-------
40
39
37
41
39
41
43
43
43
44
44
48
48
50
54
55
57
57
59
58
59
50
58
60
46
49
55
41
43
--
42
39
--
adjacent Cedar River. The data in Table 3 for Puget Sound area condi-
tions indicate that the treatment plant effluents are warmer than the
diverted water by about 14°F. in the winter, 12°F. in the spring, 9°F.
in the summer, and 13°F. in the autumn. Additional data are needed on
water temperature increase through municipal use.
Table 3 - Sewage Treatment Plant Effluent Vs.
River Temperatures, °F.
Year Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec.
Green River Intake
1961 i/ 41
1962 !/ kk
1963 !/ 36
Tacoma Sewage Treatment Plant
1961 54 54 57 59 62 64 64 67 65 62 61 54
1962 55 56 54 59 59 61 63 64 64 62 62 57
1963 57 57 58 57 60 68 67 67 65 61
Cedar River Intake
I960 39-5 40.1 40.6 44.6 48,5 52.8 56.4 54.0 51.1 48.8 43-3 ^0.3
Seattle Alki Sewage Treatment Plant I/
I960 53 53 52 55 58 60 63 64 62 61 55 55
]_/ Prior to Howard A. Hanson Dam Impoundment.
Temp, taken dai1y.
2/ After Howard A. Hanson Dam Impoundment (during period of
May to October) thermograph records,
3/ Daily thermometer reading.
Irrigation Return Flow:
The use of river water for irrigation may have beneficial as well
as detrimental effects in regard to wate«* temperature. Water is normally
stored to increase the irrigation season base flow of the river above
the points of diversion. If this water is stored in fairly deep reser-
voirs at higher altitudes, ic will result in water temperatures above
the points of diversion being lower than would prevail under summertime
conditions of natural flow. Howeve'1, downstream from points of diver-
sion where water is diverted so as to decrease the streamflow below its
-------
1.35 AC-FT/AC OF
SUBSURFACE SEEPAGE,
APRIL-SEPTEMBER
SALTS, 6)5 IBS AC
1.35 AC-FT/AC
1230 LBS/AC
SALTS APPLIED
TO SOIL
1.3 AC-FT/AC
WASTAGE,
APRIL-SEPTEMBER
SALTS, 300 LBS/AC
SALTS
TO GROUND WATER STORAGE
FOR RELEASE, OCTOBER-MARCH.
SALTS, 615 LBS/AC
2.7 AC-FT/AC
SOIL COLUMN PERCOLATION,
APRIL-SEPTEMBER
2.6 AC-FT/AC
EVAPO-TRANSPIRATION
LOSS,
APRIL-SEPTEMBER
1530 LBS/AC
FROM RIVER
IT It M
66 AC-^T/AC
DIVERTED FROM
YAKIMA RIVER,
APRIL-SEPTEMBER
Fig. 2
FATE OF DIVERTED IRRIGATION WATER
AND ITS SALTS, YAKIMA RIVER BASIN, 1959-60
13
-------
normal rate, water temperatures would be raised above those otherwise
occurring. The return of spent irrigation waters may raise or lower
the receiving stream temperature, depending upon the method of conduc-
ting this return flow.
Figure 2 will serve to illustrate the fate of water diverted for
irrigation in the Yakima River Basin which can be considered typical of
an irrigated area. Of 6.6 acre-feet per acre of diverted water in the
Apri1-September irrigation season (10), 1.3 acre-feet per acre is wast-
age resulting from over-irrigation and from canal spillage. This water
returns to the parent river in open drains during which time its tem-
perature is increased. Of the remaining 5-3 acre-feet, 2.6 acre-feet
(about 50 percent) is lost to evapotranspiration. This large evapo-
transpiration loss is responsible for the soil-cooling effect discussed
previously. About 2.7 acre-feet per acre of the previously diverted
water passes through the soil column into the groundwater stratum where
about half of it returns to the parent stream during the irrigation
season and the other half returns from bank storage during the non-
irrigation season. (The groundwater table rises during the irrigation
season and falls during the non-irrigation season.) This groundwater
seeps into open drainage channels or is conveyed to open channels via
subsurface drains. The great majority of the soil drainage in the
Yakima Valley is comprised of open drains.
Table 4 illustrates some of the water temperature changes observed
in the Yakima Valley irrigation facilities during August of 1959 and
1960. On the average, water temperature increases of 3«5°F. are expe-
rienced in 37 miles of main canal flow.
Tab!e k - Irrigation Water Temperature, Yakima Valley,
August 1959-!960 — Mean Values in °F. (10).
Diverted water to Kittitas, Roza, Wapato and Sunnyside Main Canals . .
61.5°
Water after traveling average of 37 miles in main canals .... 65.0°*
Water in sub-laterals as applied to land; average of 7 63.7°*
Water in sub-surface drains; average of 7 58.k°
Water in open drains as discharged to Yakima Rj average of 5 • • 67.0°
-'These two figures are not comparable as sampling stations
are different.
This is somewhat greater than would have been found in the river for the
same flow distance if the water had not been diverted. However, in
August, without irrigation flow augmentation, the normal river temper-
ature rise in 37 miles of flow would closely approximate this 3.5°F.
temperature rise. Water applied to the land had an average temperature
of 63.7°F. and that returned to the rive"- via open drains had an aver-
age temperature of 67.0°F., a rise of 3«3°F. Water temperature emerging
-------
from sub-surface drains, however, had an average temperature of 58.4 F.,
a drop of 5.3°?' As previously discussed, this drop in temperature is
caused by evaporation heat losses on the soil surface. Thus, if irriga-
tion water wastage can be reduced and if return flows can be conveyed
back to the parent river largely by sub-surface drains, these return
flows can be beneficial in lowering stream temperatures. During the
non-irrigation season the release of bank storage, built up during the
irrigation season, will tend to raise otherwise low water temperatures-
Figure 3 shows the seasonal temperature increase in the Yakima
River from the point of initial irrigation diversion at Easton to the
lower river at Kiona after most of the return flow has entered the
river. The figure shows an average water temperature increase during
the irrigation season of from 56.4 to 66,2°F, in the 72-mile stretch
between Parker and Kiona. This river stretch receives the majority of
the return flow and it also contains two areas of very low flows; below
the last major irrigation diversion at Parker and below the power canal
diversion at Prosser. The power canal diversion discharges back into
the Yakima River above Kiona. Figure 3 suggests that the increase in
river water temperature between Parker and Kiona, due to irrigation-
associated influences, had a maximum value of about 6-7°F. A more
rigorous study is needed, however, to validate this figure.
Impoundment Influences on Water Temperature
The impoundment of water will produce various temperature effects
on the impounded water temperature and on the downstream water temper-
ature, depending upon:
1. Volume of water impounded in relation to mean streamflow.
2. Surface area of impounded water.
3. Depth of impounded water.
4. Orientation with prevailing wind direction.
5. Shading afforded.
6. Elevation of impoundment.
7- Temperature of inflow water in relation to temperature of
impounded water.
8. Depth of water withdrawal-
s' Downstream flow rates during critical temperature period, i.e.,
an increase or decrease in flow over that occurring naturally.
15
-------
3
30
o 20
o
Q_
2
UJ
cr
UJ
\—
<
10
0
ESTIMATED NORMAL TEMP IN
ABSENCE OF IRRIGATION
86
PROBABLE INCREASE
TEMPERATURE DUE
IRRIGATION
IN
TO
IRRIGATION SEASON
68
u.
o
CL
LU
»-
o:
UJ
50
-—r
32
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
1959—1960-1961
Fig. 3
WATER TEMPERATURES AT YAKIMA RIVER STATIONS.
-------
In general, it can be said that large and deep impoundments will
decrease downstream water temperatu'es in the summer and increase them
in the winter, if withdrawal depths are low; that shallow impoundments
with large surface areas will increase downstream water temperatures
in the summer; that water periodically withdrawn from the surface of a
reservoir will increase downstream water temperatures; that a reduction
in normal streamflow below an impoundment will cause marked temperature
increases; and that "run-of-river" impoundments, when the surface area
has not been markedly increased ove< the normal river area, will produce
only small increases in downstream water temperatures.
Andrew and Geen (lM, in a study of downstream water temperatures
from the proposed Moran Dam on the Fraser River, estimated that the tem-
perature of water discharged from the reservoir through turbine and
spill outlets during the period of adult salmon migrations would range
from 43°F. on July 1 to 56°F. on September 30. These temperatures are
11°F. colder and 9°F. warmer, respectively, than average temperatures on
these dates in the undeveloped river. Churchill (7), in a study of
Tennessee Valley impoundments with deep withdrawal depths, observed a
lowering of downstream water temperarures as much as 14-15°C. below
normal summer stream temperatures.
Raphael (11) (12), in his calculations on the possible effect of
the Wanapum and Priest Rapids and the Wells and the Rocky Reach Dams on
the temperature of the Columbia River, concluded that in August, the
month of the largest temperature rise, the Wanapum and Priest Rapids
Dams would cause a temperature rise of 1 5°F• over that occurring with
natural flow. The predicted rise tor the Wells and Rocky Reach impound-
ments was 1°F. over the natural temperature rise. He further concludes
that "taking the two studies together, it can be seen that the continued
development of reservoirs on the Columbia River is doubling the temper-
ature rise of the water as it moves from headwaters to its mouth* It
seems inevitable that when the river is fully developed, the maximum tem-
perature of the water will rise even above its present high level. It
is estimated that maximum temperatures in the range of 70 to 75°F". at
Priest Rapids Dam must be considered in future planning."
Another effect of impoundments on smaller1 streams is to even out
extreme diurnal temperature fluctuations. Figure k shows diurnal tem-
perature fluctuations in the Green River1 above the Howard A. Hanson
impoundment of from 12.5-25.5°C. (5ft.5-78°F.) in the period of July 17-
2k. Figure 5 shows the even temperature discharge below the dam (and
above the Tacoma municipal water intake) where the diurnal temperature
fluctuation is about 1°F. This reduction in municipal water intake tem-
perature peaks is, of course, an advantage to the water user.
Figures 6, 7, and 8 show the temperature structure with depth in
three dissimilar reservoirs (3). In figure 6, Lake Merwin on the Lewis
River is a medium-depth reservoir •showing pronounced temperature grad-
ients in all seasons but the winie- and ea'ly spring of 1938-39- Thus,
17
-------
\1
zz
Fig. *t - Diurnal Temperature Changes on the Green River
above the Howard A. Hanson Impoundment
18
-------
Fig. 5 • Diurnal Temperature Changes on the Green River
Below the Howard A. Hanson Impoundment
19
-------
DEPTH TOT
INTAKE
TEMPERATURE OF LAKE MERWIN. 1938-39
PREWRED FROM DATA IN REFERENCE (3)
o LEWIS RIVER TEMP. UPSTREAM FROM RESERVOIR
OF GAUGING STATION ABOVE COUGAR
WATER TEMPERATURE - °C
10
Fig. 6
-------
0 '
$
CO
2!
10
20
30
UJ
UJ 1
u.
1
1
40 w ,
50
60 - o
•
A
1
i I
« •
~~~*
"\^
&£* k
™
/
I
*/
i
f
1
f
I
1
.
«r
-
"Ss
w
TEMPERATURE OF MCNARY RESERVOIR
3000 FEET ABOVE DAM - CENTER OF RESERVOIR
PREPARED FROM DATA IN REFERENCE (3) AND )
F8WS THERMOGRAPH RECORDS.
NORMAL DEPTH TO TURBINE INTAKE -65 FT.
• DOWNS
u>
UJ
•
TEMPERATURE -*C
8 10 12
JTRE
AM TEMP AT
5
0
IIMATILLA
d
4
•
14 16 16
I/
tf
J
? I
L
ll
u>
0
20
Fig. 7
-------
70
66
62
58
54
M
30
46
42
38
34
5 FT BELOW WATER SURFACE
^
50 FT BELOW WATER SURFACE-
ZEZ
100 FT BELOW WATER SURFACE
T BELOW WATER SURFACE
330 FT BELOW WATER SURFACE
hZO
18
-14
0
'2 ,
K
O
•IO
WATER TEMPERATURES-I95I
ROOSEVELT LAKE
NOTE:-TEMPERATURES TAKEN AT
SLOCK 68, BETWEEN 67
AND 69 TRASHRACKS. AT
GRAND COULEE DAM
FROM REFERENCE (3)
-6
-4
-Z
JAW. FEB. MARCH APRIL MAY JUNE JULY AUG. SEPT. OCT. NOV. DEC.
Fig. 8
-------
the water released through the turbines is colder than the upstream
water from May through September and warmer than the upstream water
from October through April (the August Lewis River temperature shown
in Fig. 6 is an anomaly). After construction of a similar reservoir
immediately upstream (Yale Reservoir), this was no longer true and the
combined effect of the two reservoirs in series (195^-55) is to pro-
duce a year-around warming of downstream water as shown in Table 6.
The lack of stratification in Figure 7 for the McNary Reservoir is due
to the shallow depth of the reservoir and the short detention period
for inflowing water. (Slight differences in temperature shown between
the downstream temperatures at Umatilla and the turbine intakes at 55
feet of depth is probably due to thermometer calibration or evaporative
cooling effects in the turbine tail race.) The sharp nonconforming
thermocline on May 19 occurred at a time when the Columbia River and
the Snake River inflows to the reservoir were about equal and the
warmer Snake River waters were contained in the upper layers of the
reservoir. By June 16, the Columbia River flow had more than doubled
that of the Snake and mixing occurred to destroy the temperature grad-
ient.
Roosevelt Lake above Grand Coulee Dam is a very long, deep and
narrow reservoir. The temperature gradient between surface and bottom
waters is small (Figure 8) except for the summer months when a differ-
ence of about 9°F. between surface and bottom was observed in late July,
1951> half of this temperature difference occurring in the upper 50
feet. Minimum temperatures were in March when the deepest water was
the warmest, this deep water being the closest to the temperature of
maximum density. Maximum temperatures for water withdrawal through the
turbines (nominal water depth of 260 feet) was in early October when
the pool level had been drawn down for irrigation and power production.
Isothermal conditions are shown for January, May, and October when
overturns are possible.
Figure 9 is a plot of Columbia River temperatures (3) at Rock
Island Dam for the mean of a five-year period before construction of
Grand Coulee Dam and a five-year period after construction of the dam.
Time periods were chosen when river flows and air temperatures were
similar. As shown in Figure 9, Grand Coulee Dam construction has pro-
duced warming of the Columbia River at Rock Island between September
and March and cooling between March and September. This warming ef-
fect was about 7°F. maximum in the winter and the maximum cooling ef-
fect in the summer was about 3°F.
Table 5 gives the characteristics of several reservoirs in
western and eastern Washington. Table 6 lists the observed tempera-
ture changes through these impoundments for different periods of ob-
servation. The Yale-Merwin impoundments in series on the Lewis River
give a significant temperature rise throughout the year. The Howard
A. Hanson flood control-conservation impoundment produced a uniform
23
-------
34
COLUMBIA RIVER AT ROCK ISLAND
TEMPERATURE COMPARISON FOR
YEARS 1934-38 AND 1946-50
PRE AND POST GRAND COULEE DAM
COMPILED FROM P.S.P.8 L. CO. RECORDS
JAN
MAR.
MAY
JULY
SEPT.
NOV.
Fig. 9
-------
temperature rise of about 1.5°F. during the period of conservation stor-
age until October, when the pool was drawn down, and the warmer surface
waters raised the flow-through temperature by 3°F. Grand Coulee im-
poundment water is pumped into the Banks irrigation equalizing pool
which is a very broad and shallow pool subject to a maximum of solar
radiation in the summer months. Temperature increases exceeded 7°F,
through this reservoir from the cold Grand Coulee inlet water to the
irrigation water discharge. In Roosevelt Lake, temperature decreases
through the reservoir are experienced in the summer until September
when the reservoir is drawn down and the warmer upper level water enters
the turbine intakes. The effect of Roosevelt Lake in cooling Columbia
River water would be more pronounced in Table 6 if the natural tempera-
ture increase through the 150-mile reservoir were considered. McNary
and Bonneville run-of-river impoundments produce very 1ittle warming
effect in the summer months, varying from 0 to 0.5°F.
Table 5 - Impoundment Characteristics
Impoundment
Yale-Merwin
H. A. Hanson
Banks
Roosevel t
McNary
Bonnevi 1 1 e
River
Lewis
Green
Col . Basin
Columbia
Columbia
Columbia
Average
Vol ume
AC - FT
XI 000
7**7
20
951
8,252
790
i*80
Average
Surface
Area
Acres
XI 000
7,3^0
0.6
2^.50
70.30
37-90
20.30
Average
Depth
Feet
101
33
39
118
21
2k
Theoro
Detention
at Average
Flow
Days
^3
10
UtO
35
2
1
Table 6 - Temperature Changes Through Impoundments -
Observed Average Monthly, °F. I/
Impoundment
Yale-Merwin
H. A. Hanson
Banks
Roosevelt
McNary I/
Bonnevi 1 1 e
Average Monthly Temperature Change Through Impoundment
Mar.
1.6+
-
-
-
1.5-
0.2+
May
3.0+
1.5+
-
.
0.1-
•
June
5.1 +
1.8+
7.0+
1.9-
0.7-
0.1-
July
1.1++
1.8+
7-5+
1.9-
0.0
0.1-
Aug.
0.8+
1.8+
5.9+
0.1-
0.1 +
0.0
Sept.
3.1 +
1.5+
2.0+
3.6+
0.5+
0.0
Oct. Nov. Dec.
U.O+
3.0+ -
. _
_
- 0.2+ 0.2-
- 0.0 0.5-
I/ From reference (3), 195% 1955; except Hanson U. of W. data,
1962, 1963.
2/ Temp, above McNary Dam measured at Pasco. Snake River inflow
raises or lowers McNary pool temp.
25
-------
Summary
A discussion has been presented on the various factors that in-
fluence water temperatures in streams, impoundments and on irrigated
lands. Data are shown that illustrate the wide range of temperature
increases and decreases that are obtained seasonally through water use
and its method of use. These data are fragmentary and indicate the
need for a thorough study on water temperature patterns as influenced
by man's alteration of the natural water environment.
Low water temperatures, commensurate with the maximum productiv-
ity of the fishery, should be the goal in water quality management for
the following reasons: Low temperatures increase the oxygen capacity
of a water body; they slow the rate of biological oxidation (which may
not always be desirable); cool water is more palatable in a municipal
system; cool water is more valuable to industry for cooling purposes;
and the rate of metal corrosion is reduced. Occasionally, cool water
is undesirable for certain crops, such as rice-
Future engineering design and redesign or operational changes in
existing structures, in consideration of the present trend towards
higher river water temperatures, can do much towards reducing or amel-
iorating this trend.
References
1. Reid, G. K., Ecology of Inland Waters and Estuaries, Reinhold Pub.
Corp., New York, 1961.
2. Velz, C. J. and Gannon, J. J., Forecasting Heat Loss in Ponds and
Streams, Jr. Water Poll. Cont. Fed., ^2, 4, April, I960.
3. Sylvester, R. 0., Water duality Studies in the Columbia River Basin,
U. S. Fish and Wildlife Serv, Sp. Scientific Report - Fisheries No.
239, May, 1958.
k. The Yearbook of Agriculture, 1955* S^th Congress, 1st Session,
House Document No. 32.
5. Sylvester, R. 0., Water dual ity Study of Wenatchee and Middle
Columbia Rivers Before Dam Construction, U. So Fish and Wildlife
Service Sp. Scientific Report - Fisheries No. 290, March, 1959.
6. Water Temperature Studies for 1958 and 1959 Middle Snake River
Drainage, Dept. of the Interior, Fish and Wildlife Service, Bur.
Comm. Fisheries, I960.
7. , Churchil1, Milo A., Effects of Storage Impoundments on Water
Qual ity, Trans. Am. Soc. Civil Eng,, 123, p. W9, 1958.
26
-------
8. Sylvester, R. 0. and Carlson, D. A., Lower Columbia River Basic
Water Quality Data Analysis for the Year I960, Um'v. of Wash.,
Dept. of CivilEngineering,Oct.,1961.
9. Hoak, R. D., The Thermal Pollution Problem, Jr. Water Poll. Cont.
Fed., 33, 12, Dec., 1961.
10. Sylvester, R. 0. and Seabloom, R. W., The Character and Signifi-
cance of Irrigation Return n_ows__in the Yakima River Basin, Univ.
of Wash., Dept. of CivilEngineering, Feb., 1962.
11. Raphael, J. M., The Effect of Wanapum and Priest Rapids Dams on
the Temperature of the Columbia River, Grant Co. P.U.O. No. 2,
Ephrata, Wash., Sept., 1961.
12. Raphael, J. M., The Effect of Wells and Rocky Reach Dams on the
Temperature of the Columbia River, Grant Co. P.U.D. No. 2,
Ephrata, Wash., Jan., 1962.
13. Rostenbach, R. E., Temperature of the Columbia River Between
Priest Rapids, Washington and Umatilla, Oregon, U. S. Atomic
Energy Comm., HW-3931*?, Unclassified, Oct., 1955.
1*t. Andrew, F. J. and Geen, G. H., Sockeye and Pink Salmon Production
in Relation to Proposed Dams in the Fraser River System, Int. Pac.
Salmon Fisheries Comm., Bulletin XI, New Westminster, B.C., Canada,
I960.
15. Feigner, K. D., An Evaluation of Temperature Reduction on Low Flow
Augmentation Requirements for Dissolved Oxygen Control, M. S.
Thesis, Oregon State University, June, 19637
27
-------
DISCUSSION
0_. How did you arrange your chlorophyll rate values? Were these on
membrane filters?
A. A cooperative group made this study—the five pulp and paper mills
in the Lower Columbia, the Washington Pollution Control Commission,
and the Oregon State Sanitary Authority. They collected the algal
samples and filtered them on membrane filters, dissolved and ex-
tracted the chlorophyll with acetone, and then ran absorbence tests.
0_. The average temperatures that were shown on the graphs—were they
taken from the means of the daily temperatures, the mean of maxi-
mums and minimums, or do they take into account the whole diurnal
cycle?
A. At Rock Island the temperature was read every six hours—midnight,
6 a.m., noon, 6 p.m. We took the mean of those temperatures.
Where we had a thermograph record, we tried to balance out the
diurnal variations so that we had the same area above as below.
In other cases where we were given only the minimum temperature
or the maximum, we took the average. We found that in most cases,
if you took the minimum daily temperature and the maximum daily
temperature and averaged the mean of that, this was very close to
working out a thermograph cycle.
28
-------
WATER TEMPERATURE REQUIREMENTS FOR MAXIMUM
PRODUCTIVITY OF SALMON
Roger E. Burrows-
Water temperatures can and do affect the productivity of salmon;
the problem, then, is that of defining the temperature limitations.
At various stages in the life cycle, productivity must be measured in
different ways, either by the number of individuals produced or by the
size of fish produced. The effect of temperature on productivity may
be measured in a similar manner. Because temperature effects are dis-
similar at different stages in the l;fe cycle, it is necessary that
these effects be discussed separately for the several stages.
The temperature of the water during the upsfeam migration and the
maturation period of the adult in rhe lake or stream affect the sur-
vival of the adult, and the water temperature at time of spawning at-
fects the survival of the eggs. High temperatures during fresh water
residence are conducive to disease development and the subsequent death
of the adult prior to spawning. Fish (19^8) cites temperatures above
60°F. as conducive to disease development in blueback salmon (Oncorhyn-
chus nerka) and temperatures above 70°F. as fatal. Royal (1953) de-
fines normal spawning temperatures of sockeye salmon (0. nerka) as be-
tween 45°F. and 55°F« and indicates that temperatures above and below
this range are conducive to reduced spawning efficiency either through
death of the adults or loss of the eggs. He concludes, "Thus water
temperature was indicated as one of the major limiting factors, if not
the exclusive limiting factor, in the timing of spawning." The Inter-
national Pacific Salmon Fisheries Commission (1962), again, reports
mortalities as high as 86 percent in sockeye runs of the Fraser River
attributed to disease occasioned by water temperatures in excess of
72°F. in certain tributaries.
Burrows (I960) defines the critical water temperature for disease
development in chinook and blueback salmon as 60°F. Above this tem-
perature, survival is dependent on the extent of injury incurred by
the fish during migration and maturation, the length of the period
between the upstream migration and spawning, the incidence of disease
organisms, and the type of holding environment.
Warm water temperatures also inhibit spawning activity. Burrows
(I960) reports the effect of unseasonably high water temperature as
delaying the spawning act.
-Bureau of Sport Fisheries and Wildlife, Salmon-Cultural Labo-
ratory, Longview, Washington.
29
-------
Brett (1958) speculates on the role of cold water temperatures as
an inhibiting factor affecting the normal endocrine balance necessary
for spawning. This speculation is confirmed by Burrows (I960) report-
ing on the effect of temperature on an exotic race of large summer
chinook salmon (0. tshawytscha) which did not spawn when water tempera
tures dropped below
The temperature at spawning not only affects the adult but the
survival of the egg as well. Combs-!/ has demonstrated that water tem-
peratures below ^2.5°F. at time of egg deposition result in progres-
sively greater mortalities until at 35°F. egg losses are practically
complete in chinook salmon and up to 50 percent in sockeye salmon.
The high temperature limitations on eggs at time of spawning have
not been clearly defined. It is extremely doubtful that the adult
could survive to spawn at water temperatures which would be immediately
lethal to the eggs.
The water temperatures most conducive to maximum productivity in
the adult salmon during its fresh-water existence range from *t2.5°F.
to 55°F. Obviously the adult can exist at temperatures beyond this
range but not under optimum conditions of survival and egg deposition.
Egg incubation temperatures affect the survival of the eggs, the
rate of development, and the size of fish produced. Combs and Burrows
(1957) and Combsl/ have defined the thresholds for normal development
for chinook salmon eggs as U2.5°F. and 57-50F. and for sockeye salmon
eggs as 42.5°F. and 55°F. In addition, Combs demonstrated that both
chinook and sockeye salmon eggs would tolerate 35°F. temperatures after
the 128-cell stage of development was reached.
Water temperatures within the thresholds of normal development
affect the growth rate of the embryo. At the higher temperatures the
growth rate is accelerated but the size of the fish produced is re-
duced primarily due to higher maintenance requirements. The size of
the emergent fry should, in theory at least, affect the survival rate
with the larger fish having the advantage. The time of emergence also
can, conceivably, affect survival particularly in species with a short
fresh-water residence. Acceleration or deceleration of emergence
could place the fish in an unfavorable environment either from the
standpoint of available forage or predation activity. Dislocation of
the time of migration could result also in unfavorable estuarine condi-
tions and poor acclimation to salt water. Vernon (1958) demonstrates
an inverse correlation between temperatures in the Fraser River during
\J "Effect of Temperature on the Development of Salmon Eggs"
Bobby 0. Combs. . Manuscript in preparation.
by
30
-------
the period of egg and fry development ot the pink salmon (0. gorbuscha)
and the size of the resultant adult run, The critical period of devel-
opment as affected by temperature appea-ed to be during hatching and
the subsequent fry stage, Decembe'- through February. Average tempera-
tures above 38°F. were usually unfavorable,- Since no temperature re-
ported was actually lethal to either eggs or fry, it must be assumed
that the acceleration of development due to the warmer temperature dis-
rupted the timing of the seaward migration.
The effect ot temperature on productivity in the egg and fry stage
of development is much more difficult to define than in any other.
While the thresholds of normal development are easy to measure, the more
obscure effects of temperature, such as its influence on fry size and
migration timing, are very evasive to evaluate but may have a profound
impact on survival.
When salmon fingeriings have a more prolonged fresh-water residence,
the effect of temperature on productivity may be measured by either the
number of. fingerlings produced or the size of finger!ings produced. The
two criteria are not synonymous because the size of the downstream mi-
grant affects survival. Marking experiments conducted with sockeye
salmon at the Leavenworth National Fish Hatchery indicated that doubl-
ing the weight of the fingeriings at release from 120 to 60 per pound
resulted in tripling the adult return. Marking experiments with fall
chinook salmon (Johnson, unpublished) indicate higher survivals for the
larger fish at release although the results are somewhat obscured by
different times of release.
Water temperature, then, to attain maximum productivity in the
fingerling must not only remain within rhe tolerance level of the
fingerling but, at least in species with more than a minimum of fresh-
water residence, reach the optimum growth level as well. Brett (1952)
defines the lethal levels for Pacific salmon fingeriings (Oncorhynchus)
as between 32°F. and 75°F. with but slight variation in the upper
threshold between species. He also found the preference temperatures
for all species to be in the range of from $0°F. to 60°F. The prefer-
ence temperatures coincide with those for optimum growth.
The response of sockeye and chinook salmon fingeriings to temper-
ature differs. Sockeye grow at a faster rate at all temperatures
between ^0°F. and 60°F. than do chinook salmon. The growth rates of
chinook salmon, however, accelerate more rapidly with temperature in-
creases. For every 10-degree rise in temperature between kO°P. and
60°F. food consumption increases *+5 percent in sockeye salmon and 60
percent in chinook salmon, but there is an initial 25 percent differ-
ence between the two species at J+0°F. While temperature may affect
species differently, it is still the prime factor in the determination
of growth rate.
31
-------
Streams or lakes in which optimum growth temperatures are not
reached, or for only short periods, will not attain maximum produc-
tivity. Similarly, streams or lakes which exceed the optimum range
for considerable periods will not attain highest productivity. Water
temperatures which remain relatively stable either above or below the
optimum range for extended periods are conducive to disease develop-
ment which may result in a reduction in the number of finger-lings
produced.
The food production capacity of a stream obviously affects fish
productivity by controlling the growth rate and time of migration.
Such food production is dependent to some extent on the temperature
regimen. Streams stocked beyond their food capacity induce forced
migrations. Chapman (1962) describes the continuous downstream dis-
placement of small silver salmon fingerlings (0. kitsutch) during the
first year of fresh-water residence and attributes this premature
migration to the aggressive behavior of the larger fingerlings.
Kalleberg (1958) suggests that the aggressive characteristic is
evolved in salmon and trout to insure an adequate food supply. It
may be concluded that food shortages result in forced migrations of
portions of the populations not necessarily at sizes and times con-
ducive to optimum survival.
The long downstream migration of salmon fingerlings imposes a
gauntlet of predation not normally encountered in other species of
fish. One of the factors affecting the degree of predation is the
activity of the predators. Water temperature controls the activity
of fish predators and, therefore, the degree of predation per fish
encountered by the downstream migrants. The preferred or temperature
of optimum activity of both the migrant and predator influences mi-
grant survival, particularly at the warmer temperatures. Streams
warm more as they progress toward the sea and the salmon migrants,
if the warm-up is considerable, may move out of their optimum activ-
ity range, thus being placed at a distinct disadvantage in the eva-
sion of resident predators.
Brett (1958), in discussing environmental stress, lists high
water temperature as an indiscriminate stress on salmon fingerlings
which may be either lethal or loading in nature. A loading stress
is defined as any environmental factor which places an undue burden
on an organism, necessitating the rapid or steady release of energy.
The warm water temperatures, between 65°F. and 75°F., encountered
on occasion during migration, place salmon fingerlings under a load-
ing stress. While such conditions may not prove immediately lethal,
they may impair the metabolic activity of the animal to such an ex-
tent that any additional stress such as pollution may prove syner-
gistic and result in a high level of mortality.
32
-------
The annual thermal cycle in a watershed determines both the produc-
tivity and the species which it will support. Brett (1959) tentatively
delineates the thermal requirements for different life processes which
characterize Pacific salmon. The work of the Salmon-Cultural Laboratory
more precisely defines the requirements in some of these areas of specu-
lation. Review of these requirements in comparison with the thermal
cycles existing in lakes and streams indicates that only rarely will
the complete thermal cycle coincide with the optimum life cycle require-
ments for maximum productivity. The thermal pattern of a stream is af-
fected by the weather conditions encountered in the area and by man-
made diversions and obstructions. Irrigation diversions can reduce the
normal streamflows until the water temperatures become intolerably
high.
Dams created for storage or power can become either liabilities or
assets depending on how they alter the normal thermal cycle of the
stream. Moffett (19^9) reports on the favorable temperature conditions
created in the Sacramento River by Shasta Dam. Johnson and Brice (1953)
describe the adverse temperature conditions created by the delayed dis-
charge of waters from the epilimnion of Dorena Dam. In the latter cir-
cumstance, the normal temperature pattern of the stream was reversed
with high water temperatures occurring in the fall during the egg incu-
bation period of the salmon.
High power and storage dams with thermal stratification provide
opportunity for control of the thermal cycle within a watercourse to
the benefit of the salmon population* Such thermal control, scientif-
ically applied, could alleviate to some extent at least some of the
detrimental effects of the dams by increasing the productivity of the
available stream area.
The effect of temperature on productivity is not confined to the
fresh-water portion of the life cycle of the salmon. Davidson (1938)
attributes limitations of the geographic distribution of the salmon in
part to the ocean temperatures encountered adjacent to the parent
stream. Tully et al. (I960) reports an intrusion of warm ocean currents
at temperatures approximating 45°F. into the Northeastern Pacific in
1957 and 1958 and attributes this intrusion as the cause for the diver-
sion of the 1958 sockeye run of the Fraser River away from its normal
migration path. Gilhousen (I960) points out that not only diversion
but delay in maturation occurred in this run and speculates that the
fish were displaced into more northern latitudes during their ocean
existence, which because of an abnormal lengthening of the light
cycle, delayed their time of maturation. Such occurrences indicate
that radical disruptions of ocean current patterns could conceivably
erect thermal barriers and completely divert a salmon population from
its normal oceanic and fresh-water habitat. Other less obvious vari-
ations in ocean temperature may have pronounced effects on salmon
productivity. Variations in the ocean survival rates of salmon
certainly exist.
33
-------
While our knowledge of the temperature effects on ocean productiv-
ity of salmon is rather nebulous at present, information on the require-
ments for the fresh-water portion of the cycle is rather precise. The
temperature requirements for maximum productivity of salmon in fresh
water may be defined as follows:
1. Temperatures during the upstream migration and maturation
period of the adult should be between U5°F. and 60°F.
2. Spawning temperatures for maximum survival of the eggs
should be between 42.5°F. and 55°F.
3. Egg and fry incubation temperatures, after 128-cell stage
of development is reached, may vary but should remain
within the range of 32°F. to 55°F. The effect of fry size
and time of migration on survival in different areas makes
it impossible to confine the optimum temperature range more
precisely.
k. The range in temperature for maximum productivity in finger-
ling salmon is between 50°F. and 60°F.
Where dams with thermal stratification make thermal manipulation
possible, every effort should be made to produce stream temperatures
compatible with optimum productivity.
References
1. Brett, J. R., Temperature Tolerance in Young Pacific Salmon, Genus
Oncorhynchus, Journal Fisheries Research Board of Canada, Vol. 9,
No. 6, pp. 265-323, 1952.
2. Brett, J. R., Implications and Assessments of Environmental Stress,
in The Investigation of Fish-Power Problems, H. R. MacMillan
Lectures in Fisheries, University of British Columbia, pp. 69-83,
1958.
1. Brett, J. R., Thermal Requirements of Fish—Three Oecades of Study,
19^0-1970, Trans. Second Seminar on Biological Problems in Water
Pollution, United States Public Health Service, Robert A. Taft
Sanitary Engineering Center, Cincinnati, Ohio, 1959.
. Burrows, Roger E., Holding Ponds for Adult Salmon, U. S. Fish and
Wildlife Service, Special Scientific Report—Fisheries 357, July,
13 pp., I960.
-------
5. Combs, Bobby D. and Roger E. Burrows, Threshold Temperatures for
the Normal Development of Chinook Salmon Eggs, U. S. Fish and
Wildlife Service, Progressive F i sh-Cul tori st, Vol , 19, No. 1,
pp. 3-6, 1957.
6. Chapman, D, E., Aggressive Behavior in Juvenile Coho Salmon as a
Cause of Emigration. Journal of Fisheries Research Board of Canada,
Vol. 19, No. 6, pp. 1C47-1080, 1962.
7. Davidson, Frederick A. and Samuel J. Hutchinson, The Geographic
Distribution and Environmental Limitations of the Pacific Salmon
TGenus Oncorhynchus), Bulletin of the U. S. Bureau of Fisheries,
Vol . 48, for 1940, pp. 667-689, 1938.
8. Fish, Frederic F., The Return of Blueback Salmon to the Columbia
River, Scientific Monthly, Vol. 46, No. k, April, pp. 283-292,
9. Gilhousen, Philip, Migratory Behavior of Adult Fraser River Sockeye,
Progress Report of International Pacific Salmon Fisheries Com-
mission, 78 pp., I960.
10. Royal, Loyd A., Annual Report for Year 1961, International Pacific
Salmon Fisheries Commission, U3 pp., 1962.
11. Johnson, Marian E. and Richard F. Brice, Use of Impounded Water for
Fish Culture, U. S. Fish and Wildlife Service, Research Report 35,
35 pp., 1953.
12. Kalleberg, H., Observations in a Stream Tank of Territorial ity and
Competition in Juvenile Salmon and Trout (Salmo salar L. and s7
TFutta L.), Institute of Freshwater Research Report 39, Drottning-
holm, pp. 55-98, 1958.
13« Moffett, James W., The First Four Years of King Salmon Maintenance
Below Shasta Dam, Sacramento River, California, California Fish
and Game, Vol. 35, No. 2, April, pp. 77-102, 19**9.
1*t. Royal, Loyd A., The Effects of Regulatory Selectivity on the
Productivity of Fraser River Sockeye, The -Canadian Fi sh-Cul turist,
October, pp. 1-12, 1953.
15. Tully, J. P., A, J. Dodimead, and S. Tabata, An Anomalous Increase
of Temperature in the Ocean Off the Pacific Coast of Canada Through
1957 and 1958, Journal Fisheries Research Board of Canada, Vol. 17,
No. 1, pp. 61-80, I960.
16. Vernon, E. H., An Examination of Factors Affecting the Abundance of
Pink Salmon in the Fraser River, Progress Report of the Inter-
national Pacific Salmon Fisheries Commission, ^9 pp., 1958.
35
-------
DISCUSSION
Q. Have fisheries agencies ever made their wants known about using
thermal structure of reservoirs for optimum production of eggs
and fry? This can be done.
A. They really did not know what was wanted for quite some time, but
they know pretty well what is wanted now.
Q.. If you don't speak up, you'll never get what you want.
A. That's for sure. I presume it is entirely feasible to adjust so
that water is taken from either the top layers or lower layers of
a reservoir without affecting the power production, if there are
outlets to do it. But in running all the water through the tur-
bines, there is not much choice.
Q. Isn't it possible, in dam construction, to provide for taking out
of different levels?
A. In new dams this can be done, and is being done now. However, in
the operation of existing dams, this is not possible. My sug-
gestion is to get letters to people who operate these dams, telling
them what your needs are.
A. It seems that it is a glaring defect in our entire fish-management
program that we have made no effort to take advantage of the thermal
requirements of the fish and to correlate them with the thermal
capabilities of management.
A. This is possible only within the limits of the overall objectives
of the project, and reservoir operation is only too happy to help
out. For instance, recently the Fish 6- Wildlife requested that
we operate the Bumping River Dams to provide for the survival of
the finger lings. This is being done.
A. We had an occasion in California, on the American River, where long
after Folsom Dam was constructed the Bureau constructed adjustable
louvers for the control of water temperature.
A. The Army Engineers and the Bureau of Reclamation, if they knew the
requirements, would at least incorporate structures into dam con-
struction and possibly even now operate dams more advantageously.
I don't think that it would be too prohibitive in cost. There are
some areas below these streams now which are apparently much more
suitable for fish than they ever were before.
36
-------
Q_. There was a report in SCIENCE a few weeks ago about -some wo--k on
temperature periodicity. Instead of regulating temperature in
constant degrees for long periods of time, the investigate1- look
the square waves of temperature in which he changed tr»e frequency
by which temperature varied up or down. THs seemed to produce
dramatic results in survival. This wa% done, I believe, with the
juveniles, but not with the adults.
A. We have found that fish adjust very readily to temperatures o* 10
to 15 degrees change.
0_. His point was that maybe the problem of temperat u-e adaptation may
be controlled by the frequency with which Jempe1 aturti are b'Oijght
up and down and the per iod at which they stay ara the'- drop
A. This is not in conflict at all with Brett's wo- k in which he r ou"d
that where he acclimated fish to a temperature, say of 65 degrees,
their lethal level of temperature would rise Where the fish were
acclimated to lower temperatures, say 50 degrees, 6 sudden rise ro
75 would probably kill the fish. Actually this would not have been
the lethal level, if the fish had had a chance to accommodate over
a considerable period of time.
0_. You mentioned better survival of the ju^en'ile^ or the larger sizes.
Do you mean survival of adults and what about the preponderance of
jack salmon?
A. This all depends on what sizes you are talk^-'g about. For example,
with silver salmon, if you increase rne size ana hold them for' a
considerable length of time, you get a preponderance of jacks back,
With fall chinook, groups have been released wner they we-e running
about 20 to the 1b. as contrasted to fish that were running about
100 to 200 to the Ib. There have been no indications of a prepon-
derance of jacks back on these fish.
0_. What is the effect of temperature on adults carrying eggs in trans-
portation water — that is, the effect of relatively high water tem-
perature on the development of eggs prior to the fish reaching the
spawning area?
A. It doesn't seem to have any effect. On the Grand Coulee, fish were
moved in hot water. The adults may die, but if the adults survive,
the temperature at which the eggs are taken is the thing that in-
fluences the egg itself.
0_. The possibility was mentioned of acclimatizing the fish to a higher
temperature regimen where they would not be quite so susceptible to
some of the things that would affect them. How long a period of
time might be involved in making such an acclimatization?
37
-------
A. This takes about three weeks, but really It doesn't amount to much
as to the adjustment. There is only about 2 degrees difference in
the lethal temperatures and lethal temperatures are not, of course,
the important temperatures anyway.
Q.. Would it affect some of the subsequent life processes that take
place where the eggs might be affected by the higher temperature?
That is, the stage where the adult is carrying the egg?
A. As far as we know, they are not affected. There are cases where
temperatures have been as high as 72 degrees and the eggs were
perfectly normal.
Q.. Mr. Burrows, you gave a list of ranges of temperatures for various
stages of life cycles of fish, and in streams such as the Rogue
River all of these stages may be taking place at one time. Is
there a range of temperature under such conditions for all of the
cycles taking place at the same time?
A. It is difficult to visualize a condition where all of them would
be taking place at the same time other than an overlap such as
when fingerlings are moving downstream while adults that have not
yet spawned are moving upstream. In this case, the temperatures
would have to come down to the range that will not affect any
particular individual part of the cycle- If the t'ish is vulner-
able throughout any stage and it is not in an optimum habitat, we
are going to have to adjust to reach that habitat. Obviously if
we have a temperature of 32 degrees that the eggs will tolerate,
when we move into a temperature with the adult and it won't de-
posit eggs at that time, or if spawning has actually taken place
and we know that this is the lethal temperature during spawning,
then ^2-1/2 degrees becomes the minimum temperature. We narrow
the limits, in other words.
38
-------
THE EFFECTS OF TEMPERATURE ON DISEASE IN FISH
E'-ling J. Ordal and Robert E. Pacha--
Fish are poiki1othermic animals; consequently they normally take
on the temperature of the water in which they are found. This fact has
a profound effect on host-pathogen relationships, since both the host,
in this case a fish, and the pathogen, the organism infecting the fish,
may be affected differently by the water temperature. Since warm-
blooded animals have a temperature regulatory mechanism which holds the
body temperature of the animal near to a particular value, it is not
feasible to evaluate specific effects of temperature on host-pathogen
relationships in such animals. The sicuation is different with fish,
since here the temperature of the host can be placed under direct ex-
perimental control.
From the literature it is evident that most fish diseases are fa-
vored by increased water temperatures. This has been our experience
with most of the diseases of fish which we have studied at the Univer-
sity of Washington. Some of the studies on this problem have been
carried out in the University of Washington Experimental Hatchery where
water temperatures are under relatively exact control. In these stud-
ies, carried out with salmonid fishes, we have found that higher water
temperatures drastically increase the effect of such diseases as kidney
disease, furunculosis, vibrio disease due to a marine vibrio, and
columnaris disease in young fish. Experience with natural outbreaks
of a number of diseases in hatcheries, as well as observations of dis-
ease in fish in natural waters, tends to confirm these findings with
most diseases.
A striking exception to the more or less general experience that
increased water temperatures favor outbreaks of diseases in fish is
found with the disease sometimes referred to as "low-temperature dis-
ease" or "cold-water disease." This disease is due to an aquatic
myxobacterium named Cytophaga psychrophila which, as its name indi-
cates, prefers low temperatures. The disease is generally found in
young silver salmon in the early spring when water temperatures are
low and in some outbreaks causes very heavy losses of young fish. As
a rule, when water temperatures increase with the annual warm-up of
the water, the disease is self-limiting and disappears.
The effects of temperature on the disease in young silver salmon
can be illustrated by an experiment carried out at the University of
''Professor, Dept. of Microbiology; Research Instructor, Dept. of
Microbiology, University of Washington, Seattle.
39
-------
Washington. A stock of young silver salmon from the Dungeness Hatchery
suffering from low-temperature disease was brought to the University of
Washington Experimental Hatchery and held at ^3°F. Each day a consid-
erable number of these fish died. After a week the remaining fish were
divided into two groups. While one group was held at ^3°F., the second
group was tempered into water at 55°. After two days, mortalities in
the.lot of fish held at 55° ceased, but deaths continued in the fish
held at J+3°F. until all the fish were lost.
On isolation of the strain of C. psychrophila in the Dungeness
fish, it was found that this bacterium was a true psychrophile, failing
to grow on culture media except when incubated well below room tempera-
ture. However, not all strains of £. psychrophila which are found on
fish behave like the Dungeness strain, since some other strains have
been found which can cause disease at higher water temperatures. In
some cases, the disease persists as water temperatures rise to 55°F.
and even higher. In the laboratory it can be shown that some strains
of C. psychrophila produce mutants which are capable of growth at higher
temperatures, and it is probable that the disease at the higher temper-
atures is due to these mutant strains.
The occurrence of mutants in some strains of £. psychrophila which
are capable of causing disease at a higher temperature illustrates the
fact that bacteria can undergo genetic change. When first discovered,
C. psychrophila caused serious disease only in young silver salmon.
Bur ing recent years new strains have appeared, identifiable as C.
psychrophila by serological methods, which have been found responsible
for generalized infections in young chinook salmon and blueback salmon.
Columnar is disease, due to another myxobacterium, Chondrgcoccus
columnar is, is now well known and recognized as a warm-water dTsease;
and a good deal of attention has been given at the University of Wash-
ington to the study of this organism and its effects on populations of
fish not only in the hatchery but also in the natural habitat.
Columnaris disease was first described by Davis in 1922 as a new
infectious disease of warm-water fishes in the Mississippi River Valley.
The disease was found in fish which had been trapped in sloughs along
the Mississippi River when flood waters receded, and which subsequently
warmed up, and in fish in a hatchery at Fairport, Iowa, at high water
temperatures. The disease was found in 15 different species of fish.
Although Davis was not able to isolate the organism causing the dis-
ease, he observed and described the organisms present in the external
lesions on the fish sufficiently well so that there is no question but
that the disease which he studied was similar to that known today.
Following Davis* report, nothing further is found in the literature
until papers by Fish and Rucker, 19^3, and Ordal and Rucker, 19M*, on
the occurrence of columnaris disease in cold-water fishes, i.e., in
salmonid fishes,-at the Fish and Wildlife Station at Leavenworth,
-------
Washington. The building of Grand Coulee Dam made it necessary to
relocate the runs of salmonid fishes which were obstructed by the dam.
In this operation adult salmon and steelhead trout were trapped at
Rock Island Dam and hauled in tank trucks to large holding ponds on
Icicle Creek at the Leavenworth Hatchery. There they were maintained
until they spawned. The fish were trapped over the period 1939 to
19^3 and were placed in the holding ponds in the period 19^0 to
At the Leavenworth Hatchery columnaris disease was first observed
in a stock of young blueback salmon in the summer of 19^2. Cultures
were isolated, and the pathogenicity of the £. columnaris was demon-
strated, thus proving that the disease in the young salmon was due to
this organism.
Subsequently, columnaris disease was found present in adult
chinook and blueback salmon, steelhead trout, white fish, squaw fish,
chub and suckers trapped at Rock Island Dam. Pure cultures of £.
columnaris were isolated from external lesions and internal organs of
some of these fish. Columnaris disease was also found in adult chinook
and blueback salmon taken in a moribund condition from the holding
ponds at Icicle Creek. It was reported by Fish and Hanavan (19^8) that
72.3 to 97»6 percent of the adult blueback salmon and 75«^ to 95.^ per-
cent of the summer chinook salmon in these ponds died before spawning
during the four-year period 19**0 through 19^3. The cause of death in
the adult salmon was not established. However, since the fish had been
hauled by tank trucks from Rock Island Dam, and since no precautions
were taken to prevent infection, it is probable that the majority of
these fish had been exposed to columnaris disease during the hauling.
Following the reports on the occurrence of columnaris disease in
salmonid fishes and the isolation of the causative agent, epidemics of
columnaris disease were reported in a number of regions of the United
States ranging from New York to the South to the Pacific Coast (Davis,
19**9» 1953). Most of the reports have dealt with the occurrence of
columnaris disease in hatcheries or'with mass mortalities in natural
impoundments, lakes or streams when water temperatures reached high
levels.
In some studies carried out in the State of Washington in postwar
years it was noted that the strains of C. columnaris isolated in dif-
ferent places varied widely in their viFulence, that is, their capa^
bility of infecting and killing young fish. A report of these studies
was published by Rucker, Earp and Ordal (1953). Thus, it was found
that strains of C. columnaris isolated from fish in the lakes and
streams and hatcheries of Western Washington and from some fish in the
Columbia River were of relatively low virulence in that they produced
a slowly progressing infection which led to extensive tissue damage
only at relatively high water temperatures. Serious epidemics due to
these strains occurred only at high water temperatures, ordinarily In
excess of 70°F. A number of hatchery outbreaks of this type were
1*1
-------
observed, and it was noted that although many fish were tost when water
temperatures ranged from 70 to 75°?., mortalities diminished or ceased
when water temperatures were reduced to 65°F, or below. Outbreaks of
columnaris disease due to this kind of strain of £. columnaris occurred
a number of times at the Samish Hatchery on Friday Creek in western
Washington. These strains were considered to be of low virulence since
they failed to infect fish unless the fish were scarified or injured by .
some means to provide a portal of entry.
In sharp contrast to these strains of low virulence, a number of
strains of C. columnar is isolated from fish in holding ponds at hatch-
eries in the Upper Columbia River Basin in postwar years exhibited
extraordinarily high virulence. Tested in the experimental hatchery
these cultures killed young salmon in less than 2k hours when fish were
exposed to dilute cultures and held at 68°F, As an illustration, an
experiment performed with a culture Isolated from a salmon at the Entiat
Hatchery in the late summer of 1951 might be cited. Twenty uninjured
chinook salmon and twenty scarified chinook salmon were exposed to a
dilute culture of C. columnaris for two minutes, then placed in a trough
of running water at 68°F. at 5:00 pnm one afternoon„ By 9tOO a.m. the
next morning 39 fish were dead, and one fish was near death* A similar
experiment performed with a cuHure isolated from fish in the Samish
Hatchery, though using smaller numbers of fish, led to no deaths in a
period of a week in uninjured fish with half of the fish which had been
scarified dying of columnaris disease in that period.
Only limited data are available on the virulence of the cultures
of C. columnaris which were isolated during the period of the Grand
CouTee fish-maintenance project, although it was "ecognized in this
early period that water temperatures played an important role in the
disease. Fish and Rucker (19^3) showed that uninjured young fish ex-
posed to a particular culture of £. columnaris for 30 minutes and held
at 70°F. died in three days. However, only 24 percent of fish which
had been exposed similarly and held at 65°F. died in the 38-day period
of the experiment. Rucker and Ordal (194*0 carried out another exper-
iment on the effect of temperature on the effect of columnaris disease
in young salmon. It was found that uninjured fish exposed to a culture
died in 72 hours when held at ~}\.60?., while 90 percent of the fish
held at 68°F., 45 percent of the fish held at 6^°f"., and 30 percent of
the fish held at 61°F. died in a week. Thus the effect of temperature
on an infection of young salmon was well recognized, although these
cultures were far less virulent than some of the cultures subsequently
studied by Rucker, Earp and Ordal (1953). The occurrence of high
virulence strains of £. columnaris in the Upper Columbia River Basin
and their apparent absence in the waters of western Washington pre-
sented an interesting problem and led to the more recent investiga-
tions..
-------
In view of the existing data on damage done to populations of
fishes by columnaris disease, the question a'ose as to the significance
of columnaris disease to the fisheries resources of the Columbia River
Basin, particularly in view of the impending construction of dams which
might serve as points of congestion where transmission of columnaris
disease might be expedited.
One reasonable hypothesis that might account for the existence of
the high virulence strains in the Upper Columbia River Basin was that
salmon became infected with ordinary strains of £, columnaris in the
Lower Columbia River, and as they migrated upstream an increase in
virulence occurred as the result of some genetic process such as muta-
tion in the strains infecting the fish,, Higher water temperatures
resulting in increased multiplication of £. columnaris would be ex-
pected to increase the likelihood of a mutation to higher virulence.
One way in which to explore the problem was to carry out field
investigations on the disease in salmon at various locations in the
Columbia River Basin and to develop methodology whereby specific strains
of Chondrococcus columnaris might be identified. Such an investigation
was begun on a small scale in 195^ and carried out through 1959- The
most difficult part of the investigation turned out to be the procure-
ment of samples of fish at various locations in the river, since this
required availability of facilities for trapping fish and permission
and cooperation of agencies such as the Army Corps of Engineers and the
State and Federal Fishery Agencies. Sampling was carried out at
Bonneville Dam, at McNary Dam, and at Rock Island Dam on the main
Columbia River; at Roza Dam, at Prosser, and at Horn Rapids on the
Yakima River; at Tumwater Dam on the Wenatchee Riverj and at Zosel Dam
on the Okanogan River. Samples were obtained at these locations when
time and the necessary cooperation and availability ot trapping facil-
ities made it possible. Since at this time there were no barriers in
the Lower Snake River, sampling was carried out in 1955 and 1956 with
the cooperation of the Fish Commission of Oregon by use of fyke nets
in conjunction with a study of patterns of migration of salmon and
steel head trout in the Snake River,
Fortunately for this investigation there were thermograph records
of water temperatures available from approximately 19^ at a number of
locations in the main Columbia River and in its major tributaries*
These records were available because of the farsightedness of Mr.
Kingsley Weber of the U. S. Fish and Wildlife Service and his associ-
ates who recognized that water temperature might be an important fac-
tor which affects runs of salmonid fishes. Unfortunately for the
investigation, the thermographs which had been employed were wearing
out, and coincident with the period of the columnaris study, most of
them were taken out of service and were not replaced. By 1958 most
of the Fish and Wildlife thermographs were out of service, and by 1959
all were removed. Though some water temperatures were taken by other
-------
agencies, all records of water temperature* used in the present study
were obtained from the Fish and Wildlife Service,
Relatively high water temperatures prevailed in the main Columbia
River during the period of the Grand Coulee Salmon Relocation Project.
This is illustrated in Figure 1, wnere mean water temperatures at Rock
Island Dam are plotted for the months of June, July, and August for the
period 1933 to 1959- Relatively warm water temperatures prevailed dur-
ing much of the 1939 to 19^3 period when upstream migrants were trapped
at Rock Island Dam. This situation was followed by a decline in water
temperatures until minimum water temperatures for July and August were
reached in 1954 and 1955- After these minimums, the temperatures again
rose, reaching a secondary maximum in 1958, and then declined rather
sharply in 1959-
temperatures in the main Columbia River in the summers of
1941 and over the period 1955 rh rough 1959 are given in Figure 2. The
water temperature at Rock Island Dam ir 1958, the warmest of recent
years, approached but remained less than that in 1941 - The Snake Rive
becomes warmer than the Columbia Rive1" during the summer months. This
river normally reaches a temperature of 65°': late in June and quickly
exceeds 70°F „ where it remains 'throughout the summer months. In 1955>
an exceptionally cold year, the warm-up was delayed approximately two
weeks. The pattern of behavior of the Snake River is given in Figure
3 where 6-day averages of daily maximum water temperatures are plotted
for the summer months. The Lower Yakima RiVer and the Okanogan River
warm up in a similar fashion, with rhe warm-up in the Yak ima River
usually occu'1'") r-g earlier '.nan in the Snake P've'".
As indicated in Figures 1 and 2, the field investigations from
1954 through 1958 covered a per >od ot f.-.c '-easing water temperatures in
the main Columbia Rive;", with a iha-p decline occurring in 1959- Par-
ticular attention was paid to blueback salmon since the patterns of
migration were known and migration occurred mainly in the latter part
of June, July and early August, T'n's pe'-iod also coincided with the
time when assistance was available f"'om students at the University.
It is not possible fo completely document the Bindings in the
period allowed for this talk. However, ^ th's location. In 1955, when
permission to sample M'sh at Bonne^'Me Dam could not be obtained,
approximately 300 scrapfish wer-e taKen *'om t he mouchs of the tribu-
taries between Bonneville and McNa'-y Dams, and only t-'ojr cultures of
£. col umnar i s could be isolated f-om these M'sh, None of these strains
exhibited high wt'ulence- In centres:, the incidence of columnaris
-------
'33 '35 '37 '39 '41 '43 '45 '47 '49 '51 '53 '55
'57 '59
Fi:
Year
1. Columbia River at Rock Island 1933-1953.
monthly ate temperature c
-------
70-
51
Fig. 2. Columbia River at Rock Island. Water temperatures
(6 day averages), 1941; 1955-1958.
1*6
-------
0^
I
§
I
iS
•sj
i
Jun.
3. Snake River at Sacajawea. Water temperatures
(6 day averages), 1955-1958.
-------
disease in fish taken in the warm waters of the Snake River near
Clarkston in July and early August of 1955 and 1956 ranged from 20
percent to 75 percent in the chinook and blueback salmon sampled by
fyke nets operated by the Fish Commission of Oregon. It was subse-
quently reported by Thompson, £t al. (1958) that 3^ percent of the
blueback salmon captured by fyke nets in the Snake River in 1955 and
50 percent of the fish captured in 1956 exhibited recognizable colum-
nar is disease.
A small run of blueback salmon enters Redfish Lake in Idaho. The
number of adult salmon reaching the weir at Redfish Lake during the
period 1955 to 1959 is shown in Table 1. It can be seen that the run
decreased from 4,361 fish in 1955 to a low of 55 fish in 1958. Evi-
dence obtained by personnel of the Oregon Fish Commission, together
with data collected in field studies carried out during this period),
suggests that the decrease in fish reaching Redfish Lake was due to
mortalities in these fish as a result of infection with columnaris
disease. By comparing the temperature data plotted in Figures 1 and
2 with the data presented in Table 1, it can be seen that the decline
in fish reaching Redfish Lake is paralleled in an inverse fashion by
water temperatures in the main Columbia River. Since the water temper-
atures of the Snake River rather consistently reached values in excess
of 70°F., it would appear from these findings that the water temper-
atures in the Columbia River below the Snake River junction played a
role in determining the incidence of columnaris disease.
Table 1
Counts of Adult Blueback Salmon Crossing
the Weir at Redfish Lake"
Year
No. Blueback Salmon
1955
4,361
1956
1,381
1957
571
1958
55
1959
290
»Data from Bjornn (I960)
In this connection, some studies carried out during the summer of
1957 are interesting. In this summer columnaris disease was essentially
absent from salmon at Bonneville Dam. When first sampled at McNary Dam
July 10 and 11, 1957, 52 blueback salmon were examined; 27 showed le-
sions characteristic of columnaris disease; 17 pure cultures were iso-
lated from these fish. Eight additional cultures were isolated from
other species of fish. Many of the lesions of the blueback salmon were
tiny and, therefore, most likely of recent origin. Hence, it seemed
probable that these fish had been infected either while in the ladders
at McNary Dam, or while massed before the dam. This might account for
the relatively high incidence of columnaris disease in the blueback
-------
salmon taken from the Snake Rive-", although the development of the
disease would, of course, be favo-ed by ihe high water temperatures
in the Snake River,
In 1955, one of the coldest years on record, samples of blueback
salmon were taken during the period of migration over Zosel Dam on the
Okanogan River- At the very first pa< the Snake River.
In 1956 only one trip wa^ made to Zosel Dam This was during the
latter part of the blueback run, and 55 percent of the fish taken were
found to be infected wi th columnaris d'sease. In 1957 columnaris dis-
ease was found common even in the ea* ly pan of the run of bl ueback
salmon. In 1958, an exceptionaHy warm yea--1, difficulties were expe-
rienced in obtaining bl ueback salmon f>om the Okanogan River during the
normal period of migration; and the r*jn was a failure. However, a
large number of blueback salmon were found dead or dying of columnaris
disease in the Similkameen River, a coole- "ributary of the Okanogan
River, where these fish had taken "'efuge.
From the studies in the Snake Ri\er and the Okanogan River it
seemed evident that: columnaris disease offered a real hazard to salmon
in the warmer tributaries, but the evidence indicated that the water
temperatures in the main Columbia River represented an important factor
in determining incidence and effects of' the disease in fish in the
tri butari es.
Beginning with the 1957 season, move attention was paid to the
question of the virulence of the suaini of £. col umnar i s isolated
from fishes of the Columbia River watershed, since it was expected
that high virulence strains would be more dangerous to salmonid fishes
than low virulence strains of C, columnar is. The facilities at the
University of Washington had been improved so that it was possible to
carry out analyses of virulence on a number of strains.
As noted before, in 1957 columnar is disease was essentially absent
from salmonid and other fishes examined at Bonneville Dam. Eight of
the 17 strains isolated from biueback salmon at McNary Dam on July 10
and 11, 1957 were analyzed for virulence immediately after isolation,
and five of these strains were found ro be of high virulence in that
they killed young salmon in less than 2k hour's when exposed to a
dilute culture in water. Sampling at Rock Island was limited to two
trips on July 16 and 2k, 1957> a* water temperatures of 61 and 6i*°F.,
respectively. In spite of these relatively low temperatures, 15 pure
cultures of £. columnaris were isolated, all from blueback salmon.
Since columnaris di'sease was not found in scrapfish at this location,
-------
the blueback salmon must have been infected while downstream. On sub-
sequent analysis for virulence, most of the Rock Island strains were
found to be of the high virulence type=
One of the tools employed for the study of columnar is disease in
fishes of the Columbia River Basin was a system of serological analy-
sis. This is a method whereby strains of a particular organism can be
distinguished from each other. The method is widely used in epidemiol-
ogical studies on certain diseases of man.
It was found that all strains of £. columnar is had a common or
species antigen. This antigen was assigned Arabic numeral _]_. Seven
additional antigens were found. These were designated by the Arabic
numerals 2, 3, 5> 6, ]^, 8, and 9. These antigens were present or ab-
sent or found present in different combinations in the strains which
were investigated.
One useful result of the development of a system of serological
analysis was that it was possible to show that wide variations in
virulence existed within a given serological type of C. columnaris.
This is illustrated in Table 2 where the virulence of~a number of dif-
ferent 1957 isolates of C. columnaris containing antigens 1, 3, 8, and
9 are compared.
Codes in this table illustrate the locations in which these cul-
tures were isolated. M is McNary; R is Rock Island; T is Tumwaterj 0
is Okanogan; and BL is Bumping Lake or Tieton Reservoir. The strains
were all isolated in 1957* and it may be noted that strains of high
virulence belonging to this serological type were isolated at a number
of locations. Interestingly enough, the strains of lowest virulence
were isolated from the Tieton Reservoir on the Naches River, a body of
water which cannot be reached by migrant fish, where a mass epidemic
of columnaris disease occurred in kokanees or silver trout at a high
water temperature. Large numbers of dead and dying fish were found
on the lake or on the shores. Tested in the laboratory, these strains
all exhibited low virulence. By present standards, this means strains
which failed to kill in four days or failed to kill at all when tested
by the contact method, that is, by exposure to dilute cultures of C.
columnaris for two minutes and subsequent holding at 68°F. An inter-
esting finding based primarily on the work in 1957 is that high viru-
lence strains occurred in many serological types. As a matter of
fact, in 1957 high virulence strains of C. columnaris were found in
seven of the serological types which were present in the Columbia
River.
The ecological and epidemiological studies on the problem of
columnaris disease in fishes of the Columbia River Basin through 1959
have been reported by Ordal and Pacha (I960) and cannot be presented
in detail here. However, it was possible to conclude that due to the
50
-------
Table 2
Variations in Virulence in Strains of £. columnari s
of Antigenic Composition 1, 3» 8, 9 and Bacter iocin Type D
Strain Virulence
(hours fo- 100% mortality in test fish)
1 -M57-22
1-M57-29
2-M57-27
3-M57-5
ii-M57-^
I-R57-2
1-&57-17
2-R57-21
2-T57-2
^-T57~7
2-057-20
3-057-37
l-BL57-le
2-BL57-3a
2-BL57-8c
2k
18
22
9V
139
22
22
20
16
1 \k
22
72
222
196
222-t-
occurrence of high virulence strains of £, columnaris in the Columbia
River Basin, columnaris disease has become one of the major factors
contributing to the decline of the >-uns of salmon and steelhead trout,
In years of high water temperatures catastrophic mortalities due to
Columnaris disease can occur in adult salmon prior to spawning. In
colder years, a larger proportion of adult salmon survive to reach the
spawning grounds, but carry high virulence strains of C. columnaris, if
present in the river, into lakes and streams supporting populations of
young salmon and steelhead trout. In light of present knowledge of
columnaris disease in the Columbia River Basin, it is probable that the
major damage to runs of salmonid fishes as a result of infection by
strains of C_. col umnar is of high virulence is not normally the killing
of adult salmon before spawning but rather the killing of young salmon
and steelhead trout in the lakes and streams supporting these fish.
Salmonid fishes such as blueback salmon, spring and summer chinook
salmon, and steelhead trout, all of which normally remain in lakes and
streams for a period of a year or more before migrating to sea are
therefore particularly susceptible to destruction by high virulence
strains of C. columnari s.
That columnaris disease was a serious threat to runs of blueback
salmon was evident from the investigations on columnaris, disease in the
51
-------
summers of 1957 and 1958. However, in the absence of counts of down-
stream migration, no data became available on damage to juvenile salmon,
As pointed out by Orda! and Pacha (I960) in a report on work supported
by the Fish and Wildlife Service, the full impact of columnaris disease
could not be determined until the return of the progeny of the 1957 and
1958 runs as adults in 1961 and 1962. That these runs were badly dam-
aged is indicated in Table 3-
Table 3
Adult Blueback Salmon Counts
at Bonnevi11e Dam
1957 to 1963
1957
1958
1959
I960
1961
1962
1963
82,915
122,389
86,560
59,713
17,111
28,179
60,027*
--'Counts up to August 1
Termination of support by the Fish and Wildlife Service for the
program of investigation of columnaris disease and its effects on the
fisheries resources of the Columbia River Basin left a number of ques-
tions unanswered.
Although it was now established that columnaris disease repre-
sented a serious threat to natural runs of salmonid fishes in the
Columbia River Basin, the question of the source of the strains of C.
columnaris, in particular, the high virulence strains, was not yet
resolved. The hypothesis that high vi-ulence strains originated from
low virulence strains in salmon infected in the lower river during the
course of migration was disproved, a^d the available evidence indicated
that the high virulence strains originated in the Columbia River Basin
somewhere between McNary and Rock Island Dams (Ordal and Pacha, I960).
Since both the Snake and the Yakima Rivers warm up early in the season,
long before the Columbia River, it was considered entirely possible
that these might be the source of the high virulence strains. Emphasis
in the earlier studies had been on salmon, but the disease had also
been found in scrapfish, particularly in scrapfish at McNary Dam during
the course of the blueback migration, and in the scrapfish observed in
the Lower Yakima River. In colder years, the disease had been found at
Rock Island Dam in salmon, but not in scrapfish. Hence, it was consid-
ered possible that the disease originated in scrapfish.
52
-------
In an effort to learn me a^'.uol sour ce o* the it -a ins of £ col um-
naris infecting fish in the Colombia Rive- Ba^i i, Meld t^ips were car-
ried out to the Yakima, Snake, and Columbia R i \ e- s early in 1962 with
support f r om t he Ai omi c Ener gy Commi ss > on 7he M - ^ t trip to the Lower
Yakima River was on May 24, 19&Z, wnen waie- t emper at ures were 59°F
On this trip 63 oui of 113 scom individual M *>h . Held trips we' e begun
to Ice Harbor Dam on the Lowe'- Snake Pi\/e' on June ] . The wate- temper -
ature at this rime was 53°F One cul'u-e was isolated, but in i he fol-
lowing week Ij cuho-es we"e isolated ai a waie- lempe* at u«-e ot 59°f'
In the following ihree weeks a la-ge p-opo-'M'or. or \ he -,;.-apf'1sh ex-
amined showed ev;dence o* coljmna'«s d'sease. 7wo cul:ufe$ OT £ Col um-
nar i s were isolated from sc'aptish ar McNaf>y Dam o^ May 3' > 1962, when
the water temperature was t)l^°f '. At McNa^y Dam d'*1 ic.lties we^e expe-
rienced in getting r i sh r o, exam;naM'on, and coOmna-. s disease was not
found to be widespread until July II, wher. 33 pu- e c^ltu-es were iso-
lated from 83 r'ibh examined Wate'' tempe'aiufes on :nis date we-e
On analysis of virulence of a numbe1" ot strains of £. col umnar j s
isolated at these sues in the earliest parr of the season, it was found
that cultures of all grades of virulence we-'e present , The presence of
low virulence strains in fish taken in the Lower Yakima River, the Lowe-
Snake River, and at McNary Dam at water temperatures ranging from 530f •
to 60°F. indicated that these strains must hav/e originated at some loca-
tion where warmer water occurred. Such a region might also provide an
environment where build-up in vi'ulence could occur through some genetic
mechanism since multiplication of bacteria infecting a po'ki lothermi c
animal would be favored by higher water lemperat u'es. Since water tem-
peratures at Rock island Dam on the main Columbia River did not exceed
52°F. over the period May 2k to Jjne ;?, 1962, during which 123 pure
cultures of C. col umnar is were i^ola'ea t • om individual fish in the
Lower Yakima River, and the water + emperar yr-es in the Lower Snake Rive-
did not exceed f>3°F, during this period, natural waters from these
streams could not have been i he wa'-me'' water in which columna* >s dis-
ease first developed.
The regions of the Columbia River fed by the warm effluents of
the Hanford reactors represent locations where initial infection of
fish by C. columnar! s and build-up of ' virulence of strains of C,
col umnar is might occur. A second possibility, suggested by personnel
of the General Electric Company, is that a number of warm sloughs on
the Hanford Reservation may be the areas where initial infection of
fish and development of high virulence strains of C ,. col umnar Is take
place.
-------
To obtain evidence on this problem, field studies were instituted
in the early spring of 1963 in order to compare cultures isolated from
fish in the sloughs on the Hanford Reservation with those from the fish
entering the Lower Yakima River. Field studies were begun in the Lower
Yakima River on April 9, 1963, at a water temperature of U8.2°F., with
the cooperation of Fish and Wildlife personnel, and continued at ap-
proximately weekly intervals. Through the cooperation of personnel of
the Atomic Energy Commission and the General Electric Company, it was
possible to obtain samples of fish, mainly by gill net, from a number
of sloughs and ponds on the Hanford Reservation. Sampling was initi-
ated on April 23, 1963, when water temperatures in the sloughs and
ponds ranged from 50°F. to 55°F.
As a result of these studies it was found that columnaris disease
was far more prevalent and cultures of C. columnaris isolated nearly a
month earlier from fish taken in the Lower Yakima River than was the
case with fish taken in the Hanford sloughs and ponds. Cultures of C.
columnaris were isolated from nine fish taken in the Lower Yakima River
on April 2k, 1963, at a water temperature of 56o3°F. Over the period
April 9 to May 27 a total of 108 cultures of C. columnaris were iso-
lated from 629 fish examined at water temperatures below 65°F. Most of
these were scrapfishes though the fish examined included some salmonid
downstream migrants. Over the period April 23 to May 16 a total of
605 fish were examined in the sloughs and ponds on the Hanford Reser-
vation. Columnaris disease was not found in these fish. Two cultures
of C. columnaris were isolated on May 22, 1963, when water temperatures
at the points of sampling ranged from 58.1°Fo to 67°F. Over the period
April 23 to June 5, 1963> when sampling was temporarily terminated by
flood waters in the Columbia River, 8^9 fish were examined and five
pure cultures of C_. col umnaris obtained.
Though analyses of virulence and of serologfcal type have not yet
been carried out, the results of the field investigations in 1963 con-
firm the findings in the spring of 1962, and it is not possible to
avoid the conclusion that the infected fish during the early season
enter the Lower Yakima River after exposure to the warm effluents of
the Hanford reactors.
At present, the evidence indicates that warm-water-loving fishes
or scrapfishes resident in the Hanford area seek the areas of warm
water in the Hanford effluents. There is some evidence from studies
in western Washington which indicates that scrapfishes carry low viru-
lence strains of £. columnaris over the winter season. Hence, when
these fish enter the warmer water these strains may develop and cause
infections in the fish. Spontaneous mutation to high virulence may
occur during multiplication in warm water, or incorporation of radio-
active materials may induce mutations to high virulence. Once infec-
ted, it is entirely possible that these fish may seek areas of colder
water, and under these circumstances multiplication of low virulence
-------
strains may cease whereas the nigh xi-ulonce strains of C. col umnan's
are capable of attacking and killing t'i ah at lowe1* temperatures. Once
high virulence sixains of C. coiumnari5 develop, there would be an
opportunity of transmission of the disease from fish to fish. Although
data on migration patterns of scrapfishes in this area are scanty, the
occurrence of columnaris disease in sc^apfishes taken in the Lower
Yakima and Snake Rivers, and subsequently at McNary Dam, can be ac-
counted for by the fact that as the water rises i'n the late spring
months the 1ish either seek the calmer waters of the Yakina or SnaUe
Rivers or are swept downstream to McNary Dam where they may accudulntr:
in the ladders and provide a source of infection of migrant salir.onid
fishes.
In conclusion, there are a numbef of problems yet to be solved.
One is the question of whether the temperature or the radioactivity oi?
the Hanford effluents is responsible fo' the development of high viru-
lence strains of C. columnaris. A second, which should be of importance
to experts in fisheries, is to determine whether high virulence strains
of C. columnari s survive from year to year in some intermediate host.
A third is consideration of the possibility that high virulence strains
of C. columnaris are transported to other watersheds by transfer of
stocks of salmonid fishes. Finally, consideration should be given to
the possibility of lowering the water temperatures of the Lower Snake
River by proper construction and operation of dams which are in the
planning stage or under consideration With construction of Lower
Monumental Dam and other dams on the Lower Snake River, there will be
multiple points of congestion where transmission of columnaris disease
to the important runs of salmonid fishes in the Salmon River will be
expedited, unless water temperatures are materially reduced during the
summer months.
-------
References
1. Bjornn, T. C., Salmon and Steel head in Idaho, Idaho Wildl ife
Review, r): 6-11, I960.
2. Davis, H. S., A New Bacterial Disease of Fresh Water Fishes,
U. S. Bur. Fisheries Bull., J§: 261-280, 1922.
3. Davis, H. S., Cytophaga Columnar is as a Cause of Fish Epidemics,
Trans. Am. Fisheries Soc., 77: 102-10^, 19^9.
**. Davis, H. S., Culture and Disease of Game Fish, Univ. of Calif.
Press, Berkeley and Los Angeles, 1953-
5« Fish, F. F. and Rucker, R. R., Columnar is as a Disease of Cold
Water Fishes, Trans. Am. Fisheries Soc., £J: 32-36, 1953.
6. Fish, F. F. and Hanavan, M. G., A Report Upon the Grand Coulee
Fish Maintenance Project of }939^Shl, U. S. Fish and Wildlife
Service Special Scientific Report No. 55,
7. Ordal, E. J. and Pacha, R. E., Final Report, Research and Investi-
gations on Diseases Affecting the Fisheries Resources of the
Columbia River Basin, I and II, 15** pp., I960.
8. Rucker, R. R., Earp, B. J. and Ordal, E. J., Infectious Diseases
of Pacific Salmon, Trans. Am. Fisheries Soc., 83: 297-312, 1953.
9. Thompson, R. N., Haas, J. B., Woodall, L. M., and Holmberg, E. K.,
Fish Commission of Oregon, Final Report, Contract DA 35-026-eng-
20609, 1958.
-------
TEMPERATURE STUDIES ON THE UMPQUA RIVER, OREGON
William H. Delay and John Seaders*
The temperature study on the Umpqua River system is an essential
part of a comprehensive study now being undertaken for the development:
of the Umpqua Basin's water resources. Central to the plan of develop-
ment envisaged by the U. S. Army, Corps of Engineers, the agency respon-
sible for the comprehensive study, are three reservoirs to be located
on the South Umpqua River, Cow Creek and Calapooya Creek. Their func-
tion is to regulate flows in each of the streams for purposes of flood
control, irrigation, municipal and industrial water supply, fish and
wildlife enhancement, recreation and water quality control.
One of the problems which needed consideration under the compre-
hensive plan of development related to stream temperature conditions.
Summer temperatures in many of the streams in the Umpqua Basin exceeded
tolerable limits for fish life and were unfavorable to the maintenance
of water quality. It was therefore recognized that a major function of
any proposed development was the control and enhancement of stream tem-
perature conditions. Capability of the proposed reservoirs 6or exer-
cising such control had therefore to be established. Information was
also needed on thermal structure in the reservoirs to enable the plan-
ning of outlet facilities for maximum utilization of the thermal jstrat-
ifications for stream temperature control. Moreover, the improvement
of stream temperature conditions had to be determined in order to fully
evaluate resulting benefits. Assessment of these benefits was consid-
ered essential for determining economic feasibility of the plan of
development.
To satisfy these planning needs, temperature evaluations were re-
quired to answer the following questions:
1. In what amounts and at what temperatures will water be
available in the reservoirs during summer months, the
season of critical stream temperature, for average as
well as critical water years and temperature years?
2. For given reservoir release rates, from what levels
should water be drawn to achieve maximum control of
downstream water temperature throughout the entire
critical period?
''•'Evaluations Engineer, State Water Resources Board, Salem;
Department of Civil Engineering, Oregon State University, Corvallis.
57
-------
3. For given release rates and temperatures, what maximum stream
temperatures will occur downstream from the reservoir, for
average as well as critical years?
A number of Federal and State agencies have been cooperating in
carrying out the temperature study, the evaluations and temperature
forecasts being the particular responsibility of the State Water
Resources Board.
For several reasons the energy-budget method was adopted for mak-
ing the necessary temperature forecasts. The method is sound theoret-
ically and basic methodology had already been developed. It had been
used by McAlister I/ for temperature predictions on the Rogue River
similar to those needed for the Umpqua River. Raphael -' had employed
the method on reservoirs and rivers in California and Washington. Al-
though there is room for refinement, the method as presently used is
expected to yield reasonably accurate results.
The method is based on the identification and evaluation of the
energy exchange processes between a body of water and its environment.
In the energy-budget equation for the body of water, all the items of
energy gain and energy loss are combined into a single algebraic ex-
pression. Solution of the equation for any given set of conditions
gives the value of energy change of the water and, hence, its tempera-
ture change.
The modified energy-budget equation as used for lakes and streams
states that for a given interval of time:
Qe = as - Qb - ae - ah + ia
where Qg = net change in energy in the body of water.
Qs = net incoming solar radiation.
Q.J, = effective back radiation from the water surface.
Q.e = energy loss due to evaporation.
Qh = energy loss by conduction from water to air.
Qa = energy advected into the water by tributary
streams, precipitation, etc.
I/ RogueRiver Basin Study by W. Bruce McAlister, 1961.
2/ Prediction of Temperatures in Rivers and Reservoirs by
~ Jerome M. Raphael, Journal of the Power Division. Proceedings
of the American Society of Civil Engineers, July, 1962.
58
-------
Although the energy-budget equation was known as far back as 1915
when Schmidt used it to estimate the annual evaporation from the oceans,
the difficulty in evaluating many of the terms has restricted its appli-
cation. That the classical energy-budget equation needed modification
was established by the exhaustive Lake Hefner Study I/ carried out in
the early fifties. This study also developed information which enabled
the evaluation of energy-exchange processes with a greater degree of
confidence. Accuracies approaching +_ percent were obtained for evap-
oration from Lake Hefner for periods greater than seven days using the
energy-budget method as compared to the water-budget method. Reference
will be made later to a verification study which was carried out on the
Willamette Coast Fork by the agencies who are participating in the
Umpqua study.
The following is a brief description of the terms of the energy-
budget equation and the data required to evaluate them:
a. Q.s - Net Solar Radiation
Between sunrise and sunset, the body of water receives energy
directly from the sun in the form of short-wave radiation.
Of the total radiation incident on the water surface, a small
portion is lost by reflection and scattering. The rest of the
energy is absorbed by the water and is identified in the en-
ergy—budget equation as Q.s, net incoming radiation. Radiation
values for Roseburg, which Is centrally located, are assumed
to be applicable for all locations in the Umpqua River Basin.
In the absence of recorded values, average monthly solar radi-
ation for Roseburg is taken from maps prepared by Sternes, ad-
justed for measured values available for Medford. For the
verification of these values, experimental radiation inte-
grators were maintained during the past summer at selected
stations. Data obtained has yet to be evaluated.
b. Q.b - Effective Back Radiation
The body of water constantly loses energy through the emission
of long-wave radiation. Energy loss is computed by applying
the Stefan-Boltzmann radiation law, using an emissivity fac-
tor of 0.97- A constant gain in energy occurs through the
absorption of long-wave radiation emitted by the atmosphere.,
Atmospheric radiation received at the water surface is also
expressed in terms of the Stetan-Boltzmann law, but includes
an atmospheric radiation factor to allow for vapor pressure
\J Water Loss Investigations: l.ake Hefner Studies, Technical
Report, Geological Survey Professional Paper 2^9,
59
-------
and cloud cover. It is estimated that only 97 percent of this
radiation is absorbed by the water.
Expressions for energy exchanges through long-wave radiation
are usually combined into one which gives the effective back
radiation. Under certain circumstances, however, the exchange
will result in a net gain in energy to the body of water.
The equation is as follows:
Q. = o.97cj~(TJ; - rj) e
where Q.^ = effective back radiation
^~ - Stefan-Boltzmann constant
Tw = absolute temperature of the water
& - atmospheric radiation factor
Tfi = absolute temperature of air
8 = time
c. Qe - Energy Loss Due to Evaporation
The evaporative process removes energy from the body of water
in the form of latent heat of vaporization, the loss being
identified by the term Qe. This loss of energy is computed
by means of an empirical equation which was found to agree
with data collected at Lake Hefner.
The equation for lakes and reservoirs is as follows:
Q,e = 0.3*» U(ew - ea) Q
j
where 0_e = energy loss in Btu/ft
U = wind speed in miles per hour
ew = vapor pressure of water in saturated air at the
temperature of the water surface, in millibars
ea = vapor pressure of water in air, in millibars
9 - time in hours
This equation also gives the energy gain due to condensation.
When applied to streams, a coefficient of 0.57 is used instead
60
-------
0.3^, to allow tor the night" -ates of evaporation from
streams.
d. Q_n - Energy Transfer by Conduction
Conduction of sensible neat occurs between the body of water
and the air whenever a temperature difference exists between
them. The rate ot conduction depends upon the temperature
differential and the wind velocity. Energy change is identi-
fied in the energy-budge* equation by the term 0_n. Values
are determined with the aid of the Bowen Ratio, which gives a
relationship between loss of energy from evaporation and the
loss from conduction. This '•atio, combined with the expres-
sion for the energy loss by evaporation, states that:
Qh -- 0.138 U(ca - tw) 6
o
where Qh = Energy transferred by conduction in Btu/ft
U - Wind in miles per hour
ta = Temperature of air in degrees Fahrenheit
tw = Temperature of wate^
-------
Data for the Umpqua River study was obtained from a number of
agencies, both from published and unpublished records. Where needed
information was not available it was secured through special studies
which were undertaken. Data was collected and processed as follows:
A. Meteorology
1. Solar Radiation
Average solar radiation for Roseburg was obtained from
Sternes radiation maps of Oregon. Values were adjusted with
the aid of Weather Bureau records for Medford, the nearest
pyrheliometer station to Roseburg. Average net daily values
corrected for reflected solar radiation were computed for 10-
day periods. One-half of the daily radiation was assumed to
occur between 0700 and 1200 hours and the other between 1200
and 1700 hours. Data was collected during the summer from
experimental radiation integrators consisting of relatively
small insulated pots containing water. Two of these inte-
grators were located at pyrheliometer stations at Medford
and Corvallis while three were located in the Umpqua Basin,
at Roseburg, Tiller and Riddle. The data collected, which
has not yet been evaluated, is expected to provide verifi-
cation of radiation values adopted for the Umpqua Basin.
2. Solar Altitude
Mean altitude of the sun was determined for each ten-day
period from declination values taken from the solar ephemeris
corrected for the latitude at Roseburg. Daily average was
obtained by multiplying the mean altitude by 0.75-
3. Sky Cover
Mean sky cover at Roseburg was obtained from Weather
Bureau records, which gave the information in tenths of sky
covered for the period between sunrise and sunset. These
values were assumed to hold for the period from midnight to
midnight. Values were average for ten-day periods.
k. Relative Humidity
Relative humidity at Roseburg was taken from Weather
Bureau records, and mean values were determined for the
periods 0000-0700, 0700-1200, 1200-1700 and 1700-2400
hours for each ten-day period.
5. Mean Winds
Mean wind speeds for Roseburg were taken from Weather
62
-------
Bureau records. Values in miles per hour were tabulated for
periods identical with those adopted for relative humidity
values.
6. Air Temperature
Using Weather Bureau records for Roseburg, average air
temperature values were determined for the hours 0000-0700,
0700-1200, 1200-1700, and 1700-2400. A set of values was
obtained for each ten-day period.
7'. Barometric Pressure
Barometric pressure for Roseburg was taken from Weather
Bureau records.
8. Evaporation
A pan evaporation station was set up at Roseburg in co-
operation with several agencies and data was collected dur-
ing the past summer.
Tables containing meteorological and other data were prepared in
a form convenient for making energy-budget computations. Table 1 is
one such table.
B. Reservoirs
1. Streamflow
Streamflow data for the reservoir sites was obtained from
U. S. Geological Survey, Water Supply Papers.
2. Pool Elevation
Depth-capacity curves prepared by the U. S. Army, Corps
of Engineers, reservoir inflow rates taken from U.S.G.S.
Water Supply Papers and reservoir release rates specified
by the Corps of Engineers enabled monthly pool elevations
to be estimated.
3. Water Surface Area
Monthly water surface areas were determined from Corps
of Engineers' area-capacity curves and estimated pool ele-
vations.
63
-------
TABIE 1
MECBOBOIDGICAL MCA
PERIOD
Month
JUNE
JULY
msjss
Day
1-10
10-21
21-30
1-10
10-21
21-31
1.10
10-21
21-31
Hour
0000-0700
0700-1200
1200-1700
1700-2400
0000-0700
0700-1200
1200-1700
1700-2400
0000-0700
0700-1200
1200-1700
1700-2400
0000-0700
0700-1200
1200-1700
1700-2400
0000-0700
0700-1200
1200-1700
1700-2400
0000-0700
0700-1200
1200-1700
1700-2400
0000-0700
0700-1200
1200-1700
1700-2400
0000-0700
0700-1200
1200-1700
1700-2400
0000-0700
0700-1200
1200-1700
1700-2400
Qs*
Net Solar
Radiation
BTU/ftS
0
990
990
0
0
1040
1040
0
0
1090
1090
0
0
1170
1170
0
0
1090
1090
0
0
1110
1110
0
0
970
970
0
0
890
890
0
0
860
860
0
/3
Atmospheric
Badiation
Ebctor
.86
.85
.86
.86
.86
.86
.86
.86
.84
.34
.84
.84
.82
.83
.83
.83
.83
.83
.83
.83
.82
.82
.81
.82
.82
.83
.83
.83
.83
.83
.83
.84
.83
.84
.83
.84
ea .
Mean Vapor
Pressure
nib
12.5
12.4
13.6
13.4
12.3
12.4
12.7
13.4
12.7
12.6
12.9
13.3
12.3
13.3
13.0
13.1
13.1
14.2
13.9
13.8
12.6
13.6
12.2
13.3
11.9
13.0
12.7
12.8
12.1
13.0
11.9
13.1
11.8
13.0
12.2
12.9
U
Mean Wind
Velocity
ntpn
3
6
10
6
3
6
11
6
3
8
12
7
2
6
12
6
3
7
12
6
3
8
14
7
3
7
12
6
2
7
12
7
3
7
12
6
*a .
Mean Air
Temperature
53
64
72
61
54
65
73
62
55
66
76
64
55
69
81
66
57
71
85
69
E7
72
87
70
54
70
83
68
54
68
82
67
53
68
80
65
* Total daily solar radiation is divided equally between the two daylight periods.
-------
*+. Inflow Temperature
Temperature of reservoir inflow was taken from records of
hydrothermographs stationed at the reservoir sites.
C. Streams
1. Streamflow
U.S.G.S. Water Supply Papers provided flow data based on
gaging records. Flow in ungaged tributaries was estimated on
the basis of drainage areas and unit yield rates.
2. Water Temperature
Stream temperature data was taken from hydrothermograph
records. Instruments were located on the principal streams in
the Umpqua Basin, the earliest installations taking place in
I960.
3« Time of Travel
Time of travel was determined in the field for three dis-
charge values for each stream except Calapooya Creek. Only
one determination was made for this stream. Discharge values
were selected for damsites to represent, as far as possible,
the range of reservoir releases adopted for the temperature
study. Travel time was measured with the aid of a tracer1
technique. A fluorescent dye, Rhodamine-B, was introduced
into a stream at a known time and point of introduction and
the time taken by the dye to reach downstream points was
observed. Sampling was done at these points with a fluoro-
meter. Stream gaging measurements were carried out simul-
taneously so that travel time was related to stream dis-
charge for every reach. A plot for each of the reaches was
then made of travel time versus discharge. From these plots
travel time val ues were taken for stream discharge rates
equal to reservoir release rates adopted for the study.
These values enabled curves to be drawn of elapsed time
versus stream miles for constant discharge rates equal to
reservoir release rates. Figure 1 illustrates curves plot-
ted for the South Umpqua River for the discharge rates of
700, 1200, and 1600 cubic feet per second.
k. Water Surface Area
Water surface areas were determined for each river mile
from a number of aerial photographs taken at different times.
With the ai'd of U.S.G.S. Water Supply Papers pertaining to
65
-------
CO
tj
0)
•H
(V
CD
&
(0
H
W
30
20
10
0
70 60 50 40 30
Miles Above Mouth
20
10
700 cfs
1200 cfs
1600 cfs
FIGURE 1 - Time of Travel of South Umpqua River
-p
(0 0)
0! 0)
-------
the time of each photograph, stream discharge values were
estimated for each river mile to correspond with each sur-
face area determination, Plots were made for each river
mile of surface area versus discharge, from these plots,
surface area values were taken off for the constant stream
discharge rates adopted for the study, Curves were then
drawn of cumulative water su-face area versus stream mile.
Figure 2 illustrates the curves prepared fo-~ the South
Umpqua River.
Reservoir Temperature
Temperature analysis of reservoirs required a knowledge of 'their
physical characteristics and method or operation. The proposed reser-
voirs were relatively deep, depths at full pool being 380, 210 and 220
feet respectively for Tiller, Galesville and Hinkle reservoirs. An-
nual regulation cycle called for evacuation in the fall to provide
storage space for flood control- Reservoirs were to be filled during
winter and spring, the period of minimum stream temperature, and were
expected to attain full pool elevations by the end of spring. During
summer, reservoirs were to be evacuated, withdrawals being made from
zones of relatively cool water for purposes of srream temperature
control.
Very little is known about changes in thermal conditions within
deep reservoirs, particularly when operated for modification and con-
trol of stream temperature. Several assumptions had, therefore, to be
made in undertaking temperature evaluations 'o' reservoirs in the
Umpqua Basin. At the beginning of summe1", each reservoir wa* assumed
to have a constant temperature throughout, with the exception of rhe
surface layer, which will be discussed presently. With the advance
of summer, the development and downward movement of thermoclines were
assumed. Validity of this assumption is borne out by records avail-
able for comparable reservoirs.
Two factors were recognized as the primary causes of reservoir
temperature changes during summer months. They were:
1. Energy exchange processes identified in the energy-budget
equation; and
2. Evacuation of water from selected elevations in the reservoir
Analysis of these two factors, therefore, formed the basis of
temperature determinations in the Umpqua reservoirs. Thermal condi-
tions were determined for each of the Summer months for various reser-
voir depths, for selected meteorologic and hydrologic conditions and
for selected withdrawaI schedules.
-------
Evaluation of the energy-budget equation required a knowledge of
reservoir surface temperature. No temperature values were available
for similar reservoirs that would have served as a guide. In arriving
at a reasonable assumption, it was argued that, because of the slow
rate of heat transfer by conduction, reservoir surfaces responded di-
rectly to meteorological conditions regardless of reservoir depths.
From the temperature standpoint, the surface layer of a reservoir was,
therefore, assumed to be analogous to a shallow pond exhibiting no
thermocline characteristics. Temperature in such a body of water is
known to be ambient, following the diurnal meteorologic cycle. Surface
temperatures of the reservoirs were, therefore, assumed to correspond
to the ambient temperatures of shallow ponds subject to the same mete-
orological conditions, This assumption was not expected to cause any
significant errors in energy-budget computations for reservoirs. In
the absence of temperature observations for shallow ponds in the Umpqua
Basin, values were taken from hydrothermograph records for Cow Creek
where thermal conditions, in the summer, were known to approximate
those governing ponds.
Gain in energy for the reservoir was determined for each ten-day
period by the energy-budget equation. Computations were made on the
basis of the four elements into which a 24-hour period was divided so
as to allow for the diurnal cycle. A sample computation is given in
Table 2. Energy gain for each month was obtained by totaling the gains
for the appropriate ten-day periods.
Reservoir gain in energy for each of the months of June, July,
August and September was then distributed within the reservoir by depth
so that resulting temperature-depth curves were comparable to curves
available for certain western Oregon reservoirs. A trial -and-error
procedure was adopted for this purpose Changes in reservoir thermal
structure resulting from the energy exchange processes at the surface
were thus determined. A typical temperature-depth curve is illustrated
in Figure 3 which shows average thermal conditions in Tiller Reservoir
at the beginning of July.
Modifications to thermal struc'ore resulting from reservoir with-
drawals were determined by a simple procedure- Water withdrawn from a
certain elevation was assumed to be ar the temperature prevailing at
that elevation at the time of withdrawal. Knowledge of the quantity
and temperature of water withdrawn during each of the summer months
permitted the correction of temperature-depth curves, which had been
prepared as described in the previous paragraph, The procedure is
illustrated in Figure U. The modified temperature-depth curve for a
given month, used in conjunction with the capacity-depth curve for
the reservoir, enables determination of tne quantity and temperature
of water available in storage for temperature control during the fol-
lowing month.
68
-------
FOE TIIJJTR KESEEVOm
Period Tuly 1 - 10
TIME
(for ten days) hours
EESEEVOIH
Volume ac. ft.
Surface Area acres
Inflow ac. ft.
Ttelease ac. ft.
TEMPERATURE
Inflow ° F
Be lease ° F
t^ (Surface) ° F
SOLAE RADIATION
Qs (Table 1.) BTU/ft2
I0WG WAVE HAPIATION
tw (Surface), " F
ta ITatle 1.) ° F
B (Table 1.1 1
Q nours
0^ BTU/ft2
EVAPORATION
U (Table 1.) umh
&w , „ m°
e^ (Table 1.) nb
6 hours 0
Qe BTU/ft~
OONWETION
U (Table 1.) aroh
ta (Table 1.) • " F
t^ (Surface) a F
6 hours
Qh BTU/ft2
EMEFGY GA1K
a -
69
69
50
0
420
61
12,180
2,500
45
-32,500
50
745
12
26.7
13.0
50
3,905
12
81
72
50
-745
420
66
14,280
2,500
45
-32,500
66
45
74
0
74
66
S3
70
2,065
C
28.G
13.1
70
3,710
C,
GG
74
70
46E
580
GG
19,720
3,500
45
-45, BOO
0000-2400
240
350,000
2,850
2,000
12,000
23,400
0,685
10,495
10
fi,210
283,000
63,000
-15C,000
93,000
190,000
THKMHi UNIT = (Joantily of heat required to raise one acre-fooi- of water 1° F.
69
-------
1300
t-4
CO
H
UJ
UJ
1200
w
1100
1000
Pool elevation
50 60
TEMPERATURE IN °F
70
FIGURE 3.
Estimated temperature gradient of Tiller Reservoir
for July 1st.
70
-------
pw
53
§
H
S
1300 —
1200
1100
1000
Change in pool elevation due to withdrawal
Temperature gradient
assuming no withdrawal
Temperature gradient
allowing for withdrawal
Depth and elevation
of layer withdrawn
50 60
TEMPERATURE IN °F
FIGURE 4. Effect of reservoir withdrawal upon temperature
gradient.
71
-------
Stream Temperature
Computational procedure for streams differed in some respects from
that used for reservoirs to allow for differences in their physical
characteristics. For a stream, it was necessary to determine maximum
and minimum daily temperatures of parcels of water from the moment they
left the reservoir to the time they reached the mouth of the stream.
Energy exchange processes had, therefore, to be evaluated for relatively
short periods so that the response of the stream to diurnaJ variations
of meteorological influences was accurately ascertained. Allowance had
to be made for stream velocity and changes in stream width since they
determined the period and areal extent of exposure of the body of water
moving downstream.
To satisfy these requirements and yet keep the number of energy-
budget computations to a minimum, the following time intervals were
adopted: 0700-1200, 1200-1700, 1700-2^00 and 2M)0-0700 hours. The end
of the second period marked the point of maximum daily temperature while
the end of the fourth coincided with the minimum,
Certain assumptions were made in evaluating energy changes for
streams. One was that thermal gradients were absent due to complete
mixing of the water. Streambeds and banks were assumed to have no in-
fluence on the water temperature. Energy contributions by thermal dis-
charges, biological and chemical processes and the conversion of kinetic
energy to thermal energy were disregarded as being minor.
Stream temperatures were determined for summer months only, the
period of critical temperature. The procedure is illustrated in Table
3 which shows the evaluation of the various terms in the energy-budget
equation for a specific case. The example deals with a hypothetical
discharge of 1200 cubic feet per second from Tiller Reservoir entering
the South Fork Umpqua River at 0700 hours. A probable gain in temper-
ature of 5-7°F. is indicated by the computation, when temperature of
water released is *t5°F. and meteorological conditions are those nor-
mally encountered during the period July 1 to 10.
With the aid of specially prepared nomographs and tables, it was
possible to expedite the procedure illustrated in Table 3« Reference
should also be made to a computer program set up at Oregon State Uni-
versity as part of this study for solving the energy-budget equation
for streams.
Table 3 indicates that the average water temperature, for the
reach of river being analyzed, has to be estimated at the very com-
mencement of the analysis. This temperature is needed for evaluation
of the terms Qk, 0^ and Qji in the energy-budget equation. In cases
where the computed average differs substantially from the estimated
value, the energy-budget computation is repeated, using a new estl-
72
-------
TABLE 3
ENERGY BUDGET COMPUTATION
FOR SOUTH UMPQUA RIVER
STREAM SECTION
Beginning River Mile 77.0
Ending River Mile (from Figure 1) 69.0
PERIOD
Month July
Day 1st - 10th
Hour 0700 - 1200
9 . 5 hrs
DISCHARGE
Q. (average for Stream Section) 1200 cfs
TEMPERATURE
Initial 1+5°F.
tw (estimated average for section) U8°F.
SOLAR RADIATION
0_s (Table 1) 1170 BTU/ft2
LONG WAVE RADIATION
tw I*8°F.
ta (Table 1) 69°F.
£ (Table 1) 83%
0 5 hrs
0_b ]k BTU/ft2
EVAPORATION
U (Table 1) 6 mph
CM 11.3 mb
ea (Table 1) 13.3 mb
8. 5 hrs
O.g Jk BTU/ft2
CONDUCTION
U (Table 1) 6 mph
ta (Table 1) 69°F.
tw **8°F.
9 5 hrs
0_h -8? BTU/ft2
TOTAL ENERGY GAIN
0_s 1170 BTU/ft2
Qb i . \k
0-e 33
0-h 87
0. 1277 BTU/ft2
TOTAL TEMPERATURE GAIN
Q 1277 BTU/ft2
A (Figure.2) 6.7 X 10$ sq ft
d 1200
9 5 hrs
At 5.7°F.
FINAL TEMPERATURE 50.7°F.
" Q.e in this case is energy gain due to condensation
73
-------
mated value. In the example shown in Table 3» the assumed average of
kQ°F. is in close agreement with the computed average of k7.8°F.
Table 3 shows further that the value of Q.t is determined firstly
for a water surface of unit area, in this instance a square foot. This
value is then applied to the surface area of the reach of stream under
analysis, 6.7 x 10° square feet in the example. The final temperature
in one reach, 50.7°F. in Table 3> is raken as the initial temperature
for the next lower reach.
In the Umpqua River studies, temperatures were computed for reser-
voir releases occurring at 0700, 1200, 1700 and 2400 hours. For each
stream, computed temperatures were plotted against river miles and
curves drawn connecting maximum and minimum points, r-espectivel y -, These
curves were taken as the limits of the diurnal temperature range for the
particular conditions considered. It is proposed to develop such curves
for three discharge values and an appropriate range of initial tempera-
tures with respect to each of the th^ee ^ese>-voi r s- Curves are to be
plotted for average as well as critical yea^s for each ten-day period
from June 1 to August 31- These evaluations have been held in abeyance
until field data collected this summer becomes available-
As mentioned earlier in the paper, a verificatfon study was under-
taken on the Willamette Coast Fork. The purpose of the study was to
determine the reliability of methodology adopted for temperature analyses
of streams. Preliminary evaluations of data gathered for the Willamette
Coast Fork indicated close agreement between computed and observed stream
temperatures during the day. At night, observed values were less than
computed values. The study demonstrated the reliability of the method,
at least for the meteorological and othe' conditions experienced during
the study.
There is considerable room fo<- improvement in the energy-budget
methodology as currently used for both reservoir and stream temperature
determinations. A simplification in computational procedure, which w'H
reduce man-hours without sacrificing accuracy, is particularly needed.
Current knowledge on evaporation from streams and on surface tempera-
tures in reservoirs appears to be inadequate t'or prec?se temperature
determinations. The need exists for an expansion in the collection and
publication of meteorological data req^i'-ed Tor energy-budget studies-
Acknowledgements
The guidance and inspiration received •from Malcolm H. Karr, Chief
Engineer of the State Water Resources Board of the State of Oregon,
under whose supervision the Umpqua Temperature Studies were carried out,
and the loyal cooperation of Robert T. Evans and B'-uce A, Tichnor, who
assisted in the Computations, are gratefully ackmowledqed.
-------
DISCUSSION
Q.. Do you feel that stream turbulence may have any effect on evapora-
tion?
A. Yes, because most of our equations have been obtained from lakes
and the lakes, in general, may be fairly placid and of a given
surface area. We feel that the river, in general, is quite turbu-
lent, having a very rough surface; in fact, most of our streams on
the West Coast are for large distances white-water rivers and we
feel that the extra surface area which is involved here will be a
considerable factor in increasing the evaporation above and beyond
what is presently being estimated.
Q.. How do you justify the use of solar radiation data from only one
point when you are considering a whole basin or a length of stream
maybe 70 or 80 miles long?
A. Solar radiation is a function of cloud cover. The solar radiation
hitting a latitude would be constant, if the sky were clear. We
felt that, in the Umpqua Basin, Roseburg is centrally located in
fhe area that we are considering, and it would provide us with an
average value that could be attainable with the degree of accuracy
that we are aiming at.
0_. Do you believe that the use of pyrhel iometers would be advantageous
in the collection of solar radiation data ro supplement the read-
ings at only one point in the basin under investigation?
A. We have had some experimental radiation integrators installed in a
number of places. They are, you might say, modified Cummings radi-
ation integrators—small insulated pors. One is located in
Corvallis alongside a pyrheliometer and another one at Medford,
also alongside a pyrheliometer, and we have a few scattered in the
Umpqua Basin. Unfortunately, we have not yet evaluated this data,
but we hope that this will give us an indication of whether our
assumption or our adopted practice of using the Roseburg values
are adequate.
75
-------
TEMPERATURE PHENOMENA AND CONTROL IN RESERVOIRS
Jerome M. Raphael*
Abstract
The variation of temperature in reservoirs with time and with dis-
tance from the surface can be predicted with confidence, using a heat-
budget approach, if proper account is taken of the varying influences
of meteorological phenomena, the geometry of the reservoir, the posi-
tion of the outlets, and the volume of the inflow and outflow. The
reservoir can be considered to be made up of a number of horizontal
layers each having the average temperature of the portion of the tem-
perature gradient in that layer. Heat moves into and out of the layer
chiefly by means or" water transported inro and out of the layer. Some
heat is transferred between layers by conduction, which is another way
of saying molecular diffusion, but this amount is physically so small
in comparison with the amount of heat moved by mass transfer that it
can safely be neglected in any engineering computation.
In the typical reservoir, water leaves the reservoir through an
outlet works or penstock. The water that leaves the reservoir is drawn
not only from the layer directly in line with the outlet, but also to
some extent from layers above and below. Any water drawn from a low
elevation must be replaced by waters from layers above, and in this
manner heat is transferred downward from the surface to lower-lying
layers. Water flowing into the reservoir at a given temperature is
considered to dive down beneath warmer and lighter layers until it
hits a colder layer, and then to mix with this layer.
At the surface itself, a tremendous amount of heat is transferred.
The primary agent here is solar -"adiation. Incoming short-wave radia-
tion is attenuated by passage through rhe atmosphere, and diminished by
clouds and reflection from the surface of the reservoir. The amount of
incoming -adiation is dependent upon the hour of the day, the day of
the year, and the altitude and latitude of the reservoir. The warmed
water surface radiates heat to outer space as long-wave radiation and
the atmosphere likewise radiates to the surface of the reservoir as
long-wave radiation. The net interchange of long-wave radiation is
termed effective back radiation and is generally net radiation to
space. Some heat is lost from the surface of the reservoir by evap-
oration, which is affected by the vapor pressure of the atmosphere,
the wind velocity and the temperature difference between water surface
""Professor of Civil Engineering, University of California,
Berkeley.
76
-------
and the air. Some heat is gained t< om ihe atmosphere by direct conduc-
tion between an- and water, And *') ra I I y , some heat may be gained
directly from rainfall.
In an engineering calculation of rhese surface temperature effects,
a finite surface layer muse be considered. Experience in swimming in
deep lakes has shown the great temperature differences between the water
at the immediate surface and that lying only a few feet beneath the su'-
face. However, it can readily be visualized that this daytime phenom-
enon must change greatly at nighttime. As air temper-ature decreases,
there will come a time when the surface skin of water is coole" than tne
air. Thus the surface itself is cooled to a temperature slightly less
than the water immediately below it. Being heavier, this water sink*
through rhe warmer water and the warmer water is in turn cooled at the
surface. Thus, there are set up local convection currents which tend to
stabilize a finite layer of water of nearly uniform density at the su--
face. In calculations made using data obtained at Shasta Rese'voi-,
temperatures computed using a 10-foot surface layer corresponded most
closely to tempe-atures measured at the reservoir. Use of a thinne'
surface layer gave Temperature fluctuations which did not seem repre-
sentative of those measured. Thus, a 10-t'oot surface laye-- has been
utilized in a number of predictions of reservoir temperature,
Time intervals for these predictions should be as short as practi-
cable if detail is needed of maximum and minimum temperatures. With 3-
hour intervals, diurnal variations are easily shown. With daily in-
tervals, detail is lost, but averages are easily obtained for gaging rhe
effect of various operating criteria on the temperature of the reserve"''
Recognition of temperature as a parameter of water quality is so
recent a phenomenon that there are only a few engineering works designed
to deliver water at a predetermined temperature. These usually involve
either a number of outlets at different elevations in a dam, or a device
that can be used to mask off all inflow except ihat from an elevation •'•>
the reservoir at the desired temperature. It must be recognized that a^
a corollary to this it is necessary to have practically continuous mon-
itoring of the variation of temperature with depth in the reservoir in
order to be able to determine the quantities of water to draw at various
elevations.
77
-------
METHOD OF COMPUTING AVERAGE RESERVOIR TEMPERATURE
Peter B. Boyer-
On Reservoir Temperatures
Introduce ion
Recent emphasis on the effect of water temperature on fish, irri-
gation, and recreation led the Portland District, Corps of Engineers to
the examination and study of reservoir and river temperatures and re-
lated data. The study is continuing, but sufficient progress has been
made to report on the analysis and results. Specifically, this report
discusses the procedure used for reproducing or determining the month -
end average reservoir temperatures. Essentially, the method consists
of solving a "heat-storage" equation, known also as "heat-balance" and
"energy-budget". The equation is employed as a guide in keeping an
inventory of the basic data, assumptions, appraisals, and computations.
The terms of the equation are defined by the ordinarily observed mete-
orologic factors. All work is conveniently arranged, symbolized, and
referenced in Table 1.
Observed Reservoir Temperatures
(Detroit and Lookout Point Projects)
Since 195^, Portland District, U, S, Corps of Engineers has been
reading thermohms, installed on the upstream face of Detroit and Lookout
Point Dams, to obtain general information on the reservoir temperature
and specifically on the temperature of water entering the powerhouse in-
take. The temperature's are recorded daily from August 1 to November 15
and weekly during the remainder of the year, between the hours of *t and
8 p.m. For this study, it is assumed that the water temperature at a
thermohm equals that at the same level in other parts of the reser-
voir. J/ The observed temperatures for 1958 are plotted in three dif-
ferent forms, shown by Figures 1, 2, and 3» The patterns are typical
of other years. Examination of the patterns reveals the following tem-
perature characteristics in Detroit and Lookout Point Reservoirs:
* U. S. Army, Corps of Engineers, Portland District, Portland.
\J Dorena Reservoir is also equipped with thermohms at 10 feet
apart. However, water temperatures were observed only during 1951
and 1952.
78
-------
TABLE 1. KE3RODUCTIOH OP 1958 MOTH-HID AVSIAGE RESERVOIR IQIPHIATOREB,. DE2BOET HWJECT, OREGON
By Heat-Storage Equation: T2 = (Si^i + I % - O T0 + 0.03 A H) /32
LUJE SYMBOL ITEM
1 Si Res. content
2 T]_ Ave. Res. temp.
3 SJL 1i Res. heat content
1* I Inflov
5 TI Av. Inflow temp.
6 I TI Heat Inflow
7 0 Outflow
8 T0 Av. outflow teup.
9 O T0 Heat outflow
10 Sky cover by clouds
11 P Possible sunshine
12 Si Incident solar rad.
13 Sr Reflect, (clear sky)
11* P Sp (11) x (13)
15 Sa Absorbed, Si - P Sr
16 Ta Air temperature
17 Tv Water surf. temp.
18 Rn Het. L.W. Rad. Loss
19 Rain: O.I or more
20 Ep Pan evaporation
21 E Res. "
22 He Evapo. heat, 1*9.5 E
23 VTa (17) - (16)
21* V Wind speed
25 EC Conv. heat, 0.89
26 H (l5)-(UB)-(22)+(25)
27 A Res. surf, area
23 .03 AH Total over res. surf.
29 ITi-OT0 (6) - (9)
30 Sg T2 (3) + (28) + (29)
31 82 Res. content
32 Tg Ave. res. temp. (30)7(31)
33 Tg Observed ave.
UNIT
SFM
op
SFM-°F
SFM
°F
SFM-°F
SFM
°F
SFM-°F
Tenths
jt
Ly/dy.
"
n
tt
°F
QF
Ly/dy.
Days
Inches.
"
Ly/day
Of
mph
Ly/dy.
'Ly/dy.
Acres
SIM-°F
SFM-OF
SFM-OF
SFM
°F
Op
JAH
1*170
to
l£6.8
1*210
to
168.1*
1*800
to
192.0
9.2
17
99
38
6
93
to
1*1
103
Ifi 1
0.3
15
1
7.1*
6
-31
220O
-2.0
-23.6
11*1.2
3570
39-6
39.6
FEE
3559
11*1.2
5130
to
205.2
3820
.to
152.8
9.2
19
150
39
7
1*5
1*5
95
IB
0.7
35
0
6.3
0
13
2280
0.9
52.1*
191*. 5
1*662
1*1.7
1*1.6
MAR
1*662
191*. 5
1765
39
68.8
1015
1*1
1*1.6
7.8
1*1
280
38
Ifi
261*
1*3
1*5
106
10 i
2.25i
1.1*
69
2
7*3
13
76
2630
6.0
27.2
227.7
5**35
1*1.9
1*1.6
Affi
5435
227 .r
3350
1*1
137-1*
2250
1*1
92.3
7-5
1*7
395
to
19
376
1*7
1*9
109
17 a
2.2
109
2
7.6
13
11*5
2970
12.9
1*5.1
285.7
6535
1*3-7
1*3-7
MAY
6535
285.7
2to5
1*7
113.0
1650
1*2
69-3
6.5
57
512
1*6
26
1*36
61
61*
127
7
i*!s
223
3
6.3
17
119
11-9
l*l*.7
31*2.3
7317
1*6.8
1*6.9
JUN
7317
31*2.3
1635
51
83.1*
151*5
1*3
66.1.
7-2
1*7
1*78
1*7
22
1*56
63
66
129
ll*
5.1*0?.
1*.2
208
3
6.3
17
102
3500
10.7
17.0
370.0
50.0
50.0
JUL
71*03
370.0
925
56
51.8
1270
1*1*
55-9
1-9
93
707
1*5
1*2
665
71
75
1W*
10.022.
8.0
396
i*
7.0
25
100
31*50
10.1*
-l*.l
376.3
701*7
53.1*
5l». 2
ADD
70U7
376-3
655
57
37-3
970
1*7
1*5-6
2.6
87
623
1*1
36
587
71
76
151
2
7-7
383
5
7-0
31
22
3320
2.2
-8.3
370.2
6720
55-1
55-7
SEP
6720
370. C
635
53
33-7
1550
50
77-5
5-9
53
372
39
21
351
61
67
11*7
8
i*!i*
216
6
7-9
1*2
-56
3070
-5-2
-1*3.8
321.2
5810
55-3
56-7
OOP
5810
321.2
755
1*8
36.2
2165
55
119.1
5-9
53
273
38
20
253
56
62
11*1
3.02i
3.5
173
6
6.1
32
-93
2630
-7-3
-82.9
231.0
1*353
53-1
52.9
HOV
1*353
231.0
1*095
1*5
51*90
50
8.3
28
139
39
ll
126
W*
51
136
18 !
0.62i
1.0
50
7
8.5
61 •
-119
2080
-7.1*
-90.2
133.1*
2895
1*6.1
1*5-3
DEC
2895
133-1*
3350
1*2
lto-7
31*55
1*3
li*8.6
9-1
19
87
35
7
80
Mt
1*5
101
^
0.5
25
l
6.6
7
-53
1800
-2.9
-7-9
122.6
2785
1*1*.0
1*3-0
RaURKS
At begin, of mo.
See line (32)
In 1OOO. See line (30)
During month
S
In 10OO. g S
During month q S
From fig. 1 or 2
In 1000.
UEWB, Salem, Ore.
From fig. 5
From n^. '> «s EH
nearly same each yr. o M
Cor. for sky cover. m
Ave. for no.
Obs'd at Detroit Dam. • H
Assumed > i3
Rn^ly-0.87 Ra; table 2
Obs'd at Detroit Dam •
i Bedford £. Detroit Dam § H
Assumed g g
Ave- for mo.
USWB, Salem, Ore. is §
Ave. for mo. 8 ss
Ave. for mo. at res. surf.
Ave. for mo.
In l(>oo.
In I'.XJO.
At end of mo.; In 1OOO.
At end of month
H ' N It II
Weighted by storage.
-------
JAN FEE
MAR
APR MAY
JUNE JULY
AUG SEPT
OCT NOV
DEC
oo
TOP OF DAM (ROADWAY)
WATER SURFACE AT TIME OF
OBSERVATION
V
x-MIN. FL. CONTROL^ POOL^
x- 4. PENSTOCK INLET
LOWER OUTLET
1200
JAN
FEB
MAR APR MAY JUNE JULY AUG SEPT
FIG. 1 WATER TEMPERATURE AND PRECIPITATION, DETROIT RESERVOIR, ORE.
NOV
DEC 1958
-------
155P
1500
CQ
*
S
*
EH
itoO
oo g
I 1350
1300
1250
MAR. 28
I
1*0
APR. 28
UPPER OUTLETS
X ^^
LOWER OUTLETS
SUMMER & SPRING
I
NOV. 15
SEPT 30
SEPT 15
OCT. 31
FALL
50 F 1^ 50 60 70 °F ^0 50
FIG. 2 OBSERVED 1958 TEMPERATURE PROFILES, DETROIT RESERVOIR, ORE.
60
70 °F
-------
JAN FEE MAR APR MAY JUNE JULY AUG SEP OCT NOV DEC
CD
MAX. W.S. EL. 1568.8, JUNE 8
MIN. W.S. EL. 1M6.5, DEC. 22
AT WATER SURFACE (ASSUMED)
AT 10 FT. BELOW WATER SURFACE
MEAN MONTHLY AIR TEMP. AT EL. 130 FT.-v
AT
-------
1. During winter, there is little or no variation of temperature
with depth and time. This is the rainy season when continuous
and complete mixing takes place from repeated filling and
emptying operations of the reservoir for flood control. Mix-
ing also takes place by wave action and by the sinking of the
surface water as it cools and increases in density.
2. The alternate warming and cooling of the reservoir surface is
a daily occurrence, particularly during summer. Estimates of
heat flux indicate that heat loss by outgoing radiation and by
evaporation generally begins to exceed heat gain from solar
radiation about two hours before sunset; the loss rate in-
creases decidedly as night approaches and continues until
about an hour after sunrise.
3- The temperature of the water below the sill of penstock inlet
varies little with time. This suggests that relatively little
of the discharge is drawn from below this level.
k. The temperature of the released water is assumed to equal that
of the water in the reservoir at a level about 10 feet above
the center line of penstock inlet.
5. Near the surface, the reservoir starts to warm in March. Sum-
mer withdrawal of cold water (in excess of inflow) and contin-
ued surface heating produce steep gradients of temperature.
The maximum surface temperatures occur late in July or early
in August.
6. Temperature stratification begins with the warming of the
reservoir surface in early March and continues through the
filling and evacuation periods.
7. By the end of August the cooling rate at the surface is fast
enough to begin reversing the temperature-depth curve near the
surface.
8. At the end of November or early in December the reservoir
attains a near-isothermal condition and remains isothermal
until March when surface warming begins again.
9. The average monthly temperature of the surface water is not
observed, but it appears to be general 1y. higher than that of
the air.
10. It is believed that streamflow entering the reservoir sinks
rapidly to a level where the existing water has the same
temperature, particularly when the velocity is reduced to
less than one foot a second.
83
-------
Heat Storage Equation
The heat content of a reservoir changes as a result of the differ-
ence in the inflow and outflow of heat energy. The increase is chiefly
due to:
1. Heat content of water entering the reservoir, and
2. Net heat gain from Solar and Sky radiation.
and the decrease in heat content results mostly from loss of heat:
1. In water leaving the reservoir,
2. By long-wave radiation, and
3. By evaporation.
In addition, a reservoir gains sensible heat by convection when its
surface is colder and loses heat when the surface is warmer than the
air.
The Heat-Storage Equation, suggested for computing the average
reservoir temperature at end of a selected time interval, is expressed
in the form:
Eq. (1) S2T2 = S]Ti + I Tf - 0 To + C A H
where $2*2 ^s t'"ie Prot*uct °f tne volume of water in the rese'-
voir and its average temperature at end of the
selected time interval;
SjTj is a similar product at the beginning of the time
interval;
I T^ is the product of the reservoir inflow and its
temperature;
0 TO is the product of the reservoir outflow and its
temperature;
C A H is the quantity of heat entering or leaving the
reservoir surface, A being the area, C a dimen-
sional constant; and
Eq, (2) H = Sa - Rn - He * HC
with H representing the net rate of heat flow across the
reservoir surface,
81*
-------
S= the absorbed solar and sky radiation,
a
R the net loss of heat by long-wave radiation,
n
He the heat loss by evaporation, and
HC the relatively small heat gain (or loss) by conduc
tion.
The following figure illustrates the terms of Eqs. (1) and (2)
*. - Bh - He± Hc) CA^ CAH
The net effect of the following heat factors is considered negli
gibly small and a reliable appraisal exceedingly difficult, if not
impossible:
1. Ground heat,
2. Bank storage and its heat content,
3. Evaporation of residual ground and surface moisture as
reservoir is drawn down,
k. Shading effect of canyon walls and trees,
5. Heating (or cooling) effect of the surrounding terrain by
long-wave radiation and by reflected short-wave radiation
reaching the reservoir surface,
6. Heating effect of possible chemical and biological changes
and density currents in the reservoir.
85
-------
Their introduction would only compound further the uncertainties
involved in the evaluation of the terms of the heat-storage equation,
with resultant decrease of confidence in the reliability of the method.
Fortunately, the cooling effect of some items on the list is compen-
sated by the heating effect of others. They are omitted from the heat-
storage equation in this analysis.
Evaluation of Terms in Eg. (1)
Table 1 includes the working data, formulas, nomenclature, com-
putations, and references to sources used as aids for evaluating the
terms of Eq. (1). The computed and observed values of month-end
average reservoir temperatures are shown on the last two lines in the
table.
As noted in Figures 1, 2, and 3» during January and February, T] =
TQ = T2 = Tf = Ta, approximately. Therefore, these are the logical
months in which computation may be started. Values of Si, S-, 0, and
A are available from an actual reservoir regulation (or from an adopted
rule curve for regulation.) Note that S2 and T2 at the end of a period
equal S] and T] at the beginning of the succeeding period.
In this verification study, the outflow temperature TQ was taken
from Figure 1 at a level 10 to 20 feet above the center of the penstock
inlet. In a hypothetical study, To is generally specified or assumed
for each period.
The average reservoir inflow temperature (Tj) for 1958 was avail-
able for the computation shown in Table 1. If Ti is not available, it
may be obtained from a graph of Tj = f (Ta), similar to that shown in
Figure k, constructed from observed river and air temperature data.
If H, the net rate of heat flow at the reservoir surface, is in
gram-calories per square centimeter (langleys) per day, the total in
30 days over the entire water surface (A acres) is
Eq. (3) 30 x ^.05 x lO? A H = 1.22 x 109 A H calories
And since ^4.11 x 10'" calories will change the temperature of one cfs-
month volume of water by 1 degree Fahrenheit, the monthly change in the
heat content of the reservoir from surface heating,
Eq. (*0 C A H = 0.03 A H cfs-month - °F.
The basic data together with the computed values of the components of
the heat f1ow H across the reservoir surface are illustrated in Figure
8.
86
-------
ROGUE RIVER. MAIN STEM
oo
50 60 70
MEAN MDNTHLY AIR TEMPERATURE, F°
MEDFORD, OREGOK
6Q
K
ROGUE RIVER AT LAUREIZURST
RIVER MILE
30 to 50 60
MEAN MONTHLY AIR TEMPERATURE, F°
PROSPECT, OREGON
FIG. k RIVER TEMPERATURE VS. AIR TEMPERATURE
-------
Eva1uation of Terms In Eg. (2)
1. Sa, the absorbed solar and sky radiation, Is determined from
Eq. (5) Sa = Sf - P Sr I/
in which Sf is the solar and sky radiation, Incident on the reser-
voir surface. It is obtained from Figure 5» which is
an adaptation of a nomograph by R. W. Hamon, L. L.
Weiss and W. T. Wilson £/.
Sr Is the reflected amount under a cloudless sky. For
a given latitude, the monthly values of Sr are nearly
the same for each year. Values of Sr for Detroit
Reservoir are found on line 13, Table 1.
P is the percentage of possible sunshine duration. In
this study P was estimated from Figure 5, using cloud
cover for a nearby Weather Bureau station.
2. Rn, the net long-wave radiation heat loss, is determined from
Eq. (6) Rn = Rw - 0.87 Ra = 1.1331 (lO'8) /~(TW + 1*60)U -
0.87 (Ta + 460/*_7
or
Rn = 60 + 6.2 T - 5.k T which is a close approximation for tem-
peratures between 30 and 85°P. In Eq. (6) Rw and Ra are the "black
body" I/ radiations at water surface temperature TW and air temperature
Ta, respectively. Values of R and 0.87 R are found in Table 2, together
with an example which shows the use of the table for estimating Rn. If
not available, TW and Ta must be estimated. In this study, T. was ob-
served at project headquarters; but Tw had to be appraised. (Throne i+/
suggests Tw = 1.05 Ta. At Lake Hefner the observed monthly averages of
Tw for the months of October through May were from 1 to 5°C. less than
Ta, and for the months of June through September they were 1 to 2OC.
more than Ta.)
\J In view of the complexity and uncertainties involved in the
appraisal of unobserved elements, one may be justified in ignoring the
reflected solar radiation term in Eq. (5), but compensating for It by
increasing the coefficient in Eq. (7) by about 10 percent. That Is,
He + P Sr - 55 E approximately.
2/ Monthly Weather Review, June 195**.
3/ R =crt^ = 8.26 (10-") t1* ly/min. P. 38, compend. of Meteor.,
1951. (t in OK).
V How to Predict Lake Cool Ing Action, by R. F. Throne, Sept.,
19517 POWER.
-------
TOO
GO
VO
PERCENT
20
COVER (SUNRISE TO SUNSET)
60 80
100
MONTH
JAN
FEE
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
DEC
PERCENT OF
0 10 20
+U
+3
-1
-2
-k
-5
-5
-U
-2
0
+2
+k
+3
+3
-1
-2
-3
-U
-4
-3
-2
:
+2
+3
+3
+2
-1
-1
-3
J*
-3
-3
-1
6
+1
+3
POSSIBLE SUNSHINE
30 Uo 50 60 70
+2
+2
-1
-1
-2
--
-3
-2
-1
0
+1
+2
+2
4-2
-1
-1
-2
-2
-2
-2
-1
:
+1
+2
+2
+1
0
-1
-2
-:
-2
-2
-1
+1
+2
+1
+1
-1
-1
-2
-2
-1
-1
0
+1
+1
+1
+1
^
0
-1
-1
-1
-1
-1
0
0
+1
90
CORRECTION TO REDUCTION FACTOR
NOTE: - TO FIND AVERAGE DAILY
INSOLATION FOR AUGUST WHEN AVE.
CLOUD COVER IS 52%, FOLLOW ARROWS
AND READ 1+63 LY/DAY. THE RE-
DUCTION FACTOR (-2) IS TAKEN FROM
TABLE.
FIG. 5 DIAGRAM FOR ESTIMATING INSOLATION IN LATITUDE ^5° N.
-------
-8 4
TABLE 2. LONG WAVE RADIATION, R « 1.1331 (10 ) (T + 460 ) , Ly/day
(Ly - Langleys = cals./cm^ =3.69 BTU/ft^)
Temp.
Op
25
26
2?
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
^3
44
Ly/day
R
624
629
634
640
645
650
655
66l
666
672
677
632
638
694
699
705
710
716
722
728
0.87 R
543
547
552
557
561
566
570
575
579
585
589
593
599
6o4
603
613
6lS
623
623
633
Tenp.
OF
45
46
46
43
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
Ly/day
R
734
739
745
751
757
673
769
775
781
787
793
800
806
812
808
825
831
837
844
850
0.87 R
639
643
643
653
659
664
669
674
679
685
690
696
701
706
712
7^
723
728
734
740
Teujp.
op
65
66
67
68
69
70
71
72
73
7*
75
76
77
78
79
80
81
82
83
84
Ly/day
R
857
863
870
877
883
890
897
904
910
917
924
931
938
945
952
959
966
973
981
988
0.87 R
746
751
757
763
768
774
780
786
791
798
804
810
816
822
828
834
840
847
853
860
Temp.
op
85
86
87
38
89
90
91
92
93
9U
95
96
97
98
99
Ly/day .
R
995
1003
1010
1017
1025
1032
1040
1047
1055
1063
1070
1078
1086
1094
1102
0.87 »
366
873
879
885
892
898
905
911
93fl
925
931
938
9^5
952
959
Example; Net Long Wave Radiation Heat Loss, Rn = RW - 0.87
Given; Tw - 60° P; Ta • 55° F
Find: Ry « 825; 0.87 Ra » 690 Ly/day from table.
RJJ - 825 - 690 = 135 Ly/day
90
-------
In Eq. (6) the coefficient 0.87 represents the monthly ratio of the
atmospheric to -black-body" radiation at Ta. Actually, this ratio varies
with loca]_ humidity and cloudiness I/ (amount, height, type, and thick-
ness), smoke and dust level of the air. West of the Cascade Ranqe, it
vanes from about 0.82 in summer to 0.8? in winter-a relatively narrow
range. The effectiveness of the incoming long-wave radiation is furthe'
reduced by the reflectivity (0.03) of water %. The reason for using
the upper limit of these ratios is to compensate for taking 1.0 as the
emissivity coefficient for water instead of 0.97, commonly used. This
modification eliminates the effort required to evaluate the local humid-
ity and cloudiness which are not ordinarily available. On thTTTubjecf,
E. R. Andersen £/ concludes that:
"Empirical relationships between atmospheric radiation and local
vapor pressure may be used if 10 percent accuracy is acceptable, pro-
vided the air mass is similar to that of the area where the original
observations were recorded. For other areas, with no consideration of
air masses involved, the accuracy of the relationships is more question.
u .' " j.' ! ° ' T° obtain more accurate methods of determining atmos-
pheric radiation, in terms of more easily available parameters, it will
be necessary to consider the total vapor content of the atmosphere as
the moisture variable, rather than the local vapor pressure."
"Without taking into account the local moisture-temperature dis-
tribution with height, we should expect nothing better than a rough
approximation of the downward atmospheric radiation in the absence of
clouds," says J. G. Charney I/.
3. _ He, the evaporation heat loss, takes place mostly during rainless
periods when surface-water temperature is within 42-80°F, In this
range, the reservoir loses approximately 1485 (varying from 1473 to
1499) calories of heat per inch depth of water evaporated from each
square centimeter. Letting E be the evaporation in inches during a
selected period of t days, the average daily rate of heat loss by evao-
oration He is 1485 (E/t). And for t = 30 days:
Eq. (7) He = 49.5 E iy/day
I/ U.S.G.S. Prof. Paper 270, Water Loss Investigation; Lake
Hefner Studies. Base Data Report. ~~ '
. y PP- 90-99, U.S.G.S. Professional Paper 269, Water Loss Investi-
gations; Lake Hefner Studies. 1954.
3/ Sec. IV, Handbook of Meteorology, 1st edition, 1945.
91
-------
In this study, the monthly values of reservoir evaporation E are
estimates, using as guides the pan evaporations observed at U. S.
Weather Bureau stations in the vicinity, the number of rainy days, and
the monthly values of lake or reservoir to pan evaporation ratios sum-
marized on page 1UO of U.S.G.S. Professional Paper 269,
Of course, such appraisal of E cannot be considered objective, but
one cannot do otherwise when required data are not available for deter-
mining with confidence the regional constants and computing evaporation
by the generally accepted Meyer formula, as Marciano and Harbeck I/ did
and derived:
Eq. (8) E, = 0.00^5 U (ew - ea) cm/day
in which U is the wind speed in mph and (e^ - ea) is the vapor pressure
difference, in millibars, between reservoir surface and air.
*t. Hc, the conduction of sensible heat, which is relatively small in
this case, was estimated from:
Eq. (9) Hc = 0.89 U (Tw - Ta) 1 y/day
derived from the Bowen ratio — '
P (tw - ta)
Eq. (10) H = B
He 1000 (^ - ea)
by substituting Ej/0. 00**5 U for (ew - ea) from Eq. (8), 0.61 for B8,
1000 for the atmospheric pressure P, (5/9) (Tw - Ta) for (ty, - ta),
and 585 E! for He.
U and TW in Eq. (9) are only appraisals, based on observed values
at a meteorologic station in the vicinity of the reservoir. TW - Ta is
the temperature difference in °F. between water surface and air, and
tw - ta in Eq. (10) is a similar difference in oc.
The working equation used as guide for estimating the monthly aver-
age flow of heat at the reservoir surface was:
H = Sj - Psr - (Rw - 0.87Ra) - ^9-5E - 0.89U (Tw - Ta) by/day
But H = Si - (60 + 6.2TW - 5.**Ta) - 55E - 0.89U (Tw - Ta)
I/ P. 67, U.S.G.S. Professional Paper 269, Water Loss Investiga-
tions? Lake Hefner Studies, 1951*.
2/ P. 104, U.S.G.S. Professional Paper 269, Water Loss Investiga-
tions, Lake Hefner Studies, 1951*-
92
-------
was found to be a satisfactory app^oximati on. The latter may be used
when one cannot evaluate with confidence the reflected solar radiation
and the emissivity of the air, Furthermore, the latter is also a more
suitable form for a digital compute1".
Appli cat ion
Following the verification step described in this report, the hear
storage equation was applied for estimating the month-end average tem-
perature of water in a proposed reservoir during a critical year of low
runoff and above-normal summer-air temperature. A chart showing the
temperature variation with depth and time, similar to Figure 1, was
drawn in such a manner that the weighted average temperature at the
of each month equalled the computed month-end reservoir temperature.
Of course, the assumed monthly average reservoir-surface temperatures
and the adopted 'design temperatures of the outflow were also employed
as guides in drawing the isotherms.
Future Work
As time permits, efforts will be made to simplify the procedure, to
increase the reliability or confidence in the appraisal of unavailable
but necessary factors, and shorten the unit time from 30 to 10 days o"
less. Exploratory statistical analysis is under way to find a relax io--'-
ship between the heat content of a layer of water in the reservoir and
the pool elevation at the end of the month. Such a relationship, if
satisfactory, will serve to distribute the computed month-end average
temperature throughout the depth of the reservoir with more confidence
Water surface temperature, which is not available, makes the task
of evaluating long-wave, evaporation, and convection'heat terms of Eq
(2) very difficult. Plans a«~e ready to instrument Lookout Point,
Detroit and Fern Ridge Reservoirs for the purpose of assembling suffi-
cient surface-water temperatures which can be studied in relation to
the solar-radiation estimates or to air temperature and land pan evap-
oration, ordinarily observed at the project.
Conclusion
This is an office progress report on the analysis of Lookout Point
and Detroit Reservoir water temperatures collected since 195^ by the
Portland Office of the U. S. Corps of Engineers. The month-end average
reservoir temperatures have been reproduced with acceptable reliability,
using the heat-storage equation as a guide, and the procedure was ap-
plied in estimating the temperature pattern in several proposed reser-
voirs during critical years of low runoff and above-normal summer tem-
perature.
93
-------
The most difficult task is the numerical evaluation of the heat
flow across the reservoir surface because of the uncertainties in-
volved in the conversion of the ordinarily observed meteorologic ele-
ments at some Weather Bureau station to those over the reservoir.
Considerable judgment and trial-error method must be exercised In the
appraisal of the heat terms in Eq. (2). This task will continue to
be difficult until the reservoir surface temperature, evaporation,
distribution of ,the local humidity, wind and air temperature are in-
dexed to a satisfactory degree of approximation to the ordinarily ob-
served elements at a Weather Bureau station and the short-wave and
long-wave radiation can be determined with confidence for cloudy sky
conditions.
As noted in Figure 6, the relative magnitudes of the inflow and
outflow have greater influence on the average temperature of the water
in Detroit Reservoir than the heat flow across the reservoir surface.
Near the surface, however, the change in water temperature is almost
entirely due to heat exchange taking place at the water surface-air
interface.
The recent emphasis on water quality demands continued effort to
close the gaps in observational data and analysis leading to the sim-
plest procedure possible for satisfactory estimates of reservoir and
river temperatures.
Acknowledgment
The aid given by C. Pedersen, Chief, Water Control Section, in
reviewing this paper, is gratefully acknowledged. The assistance of
Orville Johnson (Hydr. Engr.) in collecting, processing, and charting
the basic data is also appreciated. Reservoir temperature data are
collected by project personnel under the supervision of Donald Heym
and Donald Westrick, the Project Engineers for Lookout Point and
Detroit Reservoirs.
-------
1400. £_ M J J A S 0 _N D.
I I ~T~ ~T~ ~T~ ~T~ "~T~ "T~ ~~T~ ~~T~
300
[Ti - OT0 + 0.03 AH
200
.100
-2.O
INFLOW IIEAT,
SURFACE HEAT, J . i H AH
- S.Z -7.3 -7.4
1
-100
-200
-300
^-OUTFLOW IfEAT, OT(
-UOO
-I 1 1 1 '' I I
J L
•' MAMJJAS
1958
FIG. 6 RELATIVE MAGNITUDES OF HEAT TERNE IN EQ. (l)
0 N D
95
-------
6o
I i
55
50
Observed
Compu-ted
35
Source: Ta~ble 1
I I
M
M
N
1958
FIG. 7 COMPUTED & OBSERVED MDNTH-END AVERAGE
RESERVOIR TEMPERATURES
96
-------
. J
K-
•C.
^
"*^
K'
"7 «4- —
UJ
0
O
^* i?
0
a:
UJ
tf\
t ^ 1
^•^
UJ
ft:
QC
r
K
d.
O
uJ
uJ 5
to
f
f f
«>
.
1
---
• -,
y
<
I
f H
i A
f>
^
—
1
i
Wfc
4
C
•M
e.
4
j \
r ^|
^
A/
• -
C
/r
—
i
o
W
e
#
" f ~
— 1
_ j
H
i J
f
j
i
j-
i —
!i
>^
....
T
,
1
, -M
V
J
ij;
+
^
<
i
ii
4
< 4
^
r?
. ,
/
^
1
//
I
•
:J
3
r-
^
;
5
-
-1
'<7
'
•
(1
-^
5
4
^
/
LJ
j
^
^
4
M n a
•
5
/
V
(
G
i
!
T
r-
i
1
r
i
]
^
i
i
i
'
4
t
»
>
.
i
I
n
1
,
1
f
i
:
o
A
4
,
1
'
(
I
{
i
\
1
-i
«C
^
i
5
^>
•
*
*•
^
i
\
•x^
N
j\
t
)
S
X
(
1
V
• 1
\ei
\
j
3
«
!S
•
t
V
/>
J.
.
s
1
>
n
•
Q
v
I
(.
K
J
5
5
1
^
-
^
-
^
jj
<
t
-».
i"
x
i
i
4
<
|
/
-'
0
I
7
5
,\
c
F
i
-N
s
v<
.',
/
/
r>
J
V
/
r
N
J
-)
-
/
/:
T7.
•
^
/<
K
i
\.
A
y-
^-S
.
r
•
/
>
!
1
/
?,
C?
-f
b
7
51
f
°
T
5
-.
5
9
^
;
/v
p
\
n,
S
c
C
n
j
«
- 4
;
^>
/c
'C
?A
i
r/
( v
7
• "•
••«
c^;
-/
/
}A
/^
i
J
^
P
y<|
r
<
/
*
,
•-
•),
^
»
a
f
'
I
~-
i
—
7
6
,c
,-
/
?
f-
fl
5
i
7
d
l^
^
^
^
^
9
/=
?
7
f
1
-f
|
,-
?
?/
B.
V
/=
f-1
T
<
—
/
1^
?
r
-
<
^
(
i
t-
7
9
y i
r
?
/
r^~,
M
r
3
z
/
/
-
•
c
•~
t
•
f-
•7
h
-
J
i
-
(
L
^
^
*r-
-T
C
^
'/
la
f
\i\
I
-,•
f
^c
0
.
?^
d
,
3
F
>;
F
Vi
r
•
c
,«
1
?/
^
3^
?/
^-
K
>
s
/
I
1
£
3<
w
.
;
\
^
„
f?
j
y
j
f
i
&
?>
*\
/
i
•
4
lv
\
(
'-
-
•
/
f
A
n
-
\
s
•
A
—
f
f~
*
v
>
\
\
•
\
-
f
)
'/
—
r
/
- -
k
v\
^
Ai
•k
1
C>
?2
£/
"•jr
*,
3*
7
__
5?
•
'J3
i'
97
-------
SOME OBSERVATIONS OF COLUMBIA RIVER AND RESERVOIR BEHAVIOR
FROM HANFORD EXPERIENCE
R. T. Jaske*
Considerable interest in reservoir behavior has been generated as
a result of the desire to extend the purposes of impoundments to include
regulation of downstream water quality. Churchill ]_/ on a previous oc-
casion has stated, based on his 26 years' experience, "It can be con-
cluded that water control structures may exert a profound influence dur-
ing warmer months of the year on water temperature of both impounded and
released waters. By understanding the forces that control these influ-
ences, advantage can be taken of the desirable effects, and the less de-
sirable effects can be controlled or perhaps avoided." In our own expe-
rience with Lake Roosevelt and the Columbia River we have come to an
identical conclusion, although from an entirely different set of circum-
stances.
Where Churchill and other investigators have dealt with large im-
poundments on streams of relatively low flow, the Columbia River system
of dams, with the exception of Grand Coulee, is a system where daily
through-put is a significant fraction of reservoir capacity. In some
projects, such as Rocky Reach, Rock Island and Priest Rapids, this could
involve as much as 30 percent or more of the available storage. As a
result, despite heads ranging to 90 feet, we find little effective strat-
ification. Rather, the stream has been slowed and subjected to in-
creased exposure to solar radiation, heat transfer and bank flow effects
with the net effect of a persistent increase in temperature over the
natural conditions. Attempts to explain these effects have not fully
yielded to rational explanation although Raphael £/ has pointed a way
evolved from earlier work by Anderson, et. al., at Lake Hefner and Lake
Mead. Our measurements during 1962 and 1963 fail to confirm the values
predicted in the reports, although the general method appears to have
merit. Rather, it appears that Raphael's correlation requires the addi-
tional benefit of a broader meteorological data base and machine compu-
tation from an improved mathematical model.
"Principal Engineer, Facilities Engineering Section, Irradiation
Processing Department, General Electric Company, Richland, Wash.
I/ Symposium on Streamflow Regulation for Quality Control held in
Cincinnati, Ohio, April 3-5, 1963.
2/ The Effect of Wanapum and Priest Rapids Dams on the Temperature
of the Columbia River, September,1961.
98
-------
The behavior of Lake Roosevelt mo'-e closely resembles the eastern
situation with the exception that the relatively high through-out and
low BOD demand of tributaries currently results in little or no effect
on dissolved oxygen. Lake Roosevelt does, however, develop considerable
stratification, as much as 7 degrees to 9 degrees C. during August and
September. Temperature soundings show a characteristic slope of iso-
therms downward toward the 1030-foot discharge. Correlation of the data
from various stations suggests a discontinuity rising approximately ac-
cording to the one-third power of the distance from the outlets. Fu--the
measurements in the unusual year of 1963 indicate that as of October 2f>,
the cooler inflow has failed to appear in the turbine discharge at the
rate estimated by displacement of withdrawn water, the net result being
that discharges from Grand Coulee have remained relatively constant
through September and October Ic might appear that, due to the dO>e--
ence in weather over the lake length and the sharp curvature of the old
stream bed, the lake is filling by displacement from The north rathe'
than the expected layer flow. The data will be published by the autho-
at a later date.
Additional work by the Irradiation Processing Department in suppo-:
of the river-cooling program will include attempts to derive a mathe-
matical model with computation in the JBM 7090 machine at Hanfordo At
present, we are collecting the following data for the passive record:
1. Continuous monitoring of temperatures at Grand Coulee Dam at
levels of -5 feet, -20 feet referred ro surface at the penstock
level approximately 230 feet deep.
2. Once-a-shift readings of Chief Joseph turbine discharge.
3. Hourly turbine discharge temperatures at Rocky Reach,,
k. Continuous temperature monitor ing at the Priest Rapids gage
We expect to add a continuous recorder to a point downstream ot
Grand Coulee by the first of April, 196*4. To the extent of available
resources, we expect to continue temperature soundings of Lake Roosevelt,
but these will remain somewhat fragmentary.
Downstream from the plant, the Hanford Laboratories conduct a re-
search and development program which includes an investigation of the
effects of reactor effluent on Columbia River water quality. The pur-
pose of this study is to distinguish any net changes in river water
characteristics due to plant operations from those which would occur
naturally. The emphasis is on water temperature variations, but poten-
tial chemical effects are also being studied. The work is in addition
to our routine and special studies of radioactivity in the river,.
The various phases of this work include the following:
99
-------
a. Continuous upstream and downstream monitoring at several points,
plus repeated comparisons of parameter distribution in cross
sections with the continuous point monitor data.
b. Dye studies to define diversion patterns from individual release
points, as well as the labeling of water masses in order to fol-
low time sequential changes of temperature.
c. Measurement of pertinent variables for a calculated heat budget,
using portable meteorological instruments. The portability
feature permits comparison measurements along the river with
routine data from our Hanford meteorological station.
In summary, we believe that continued study of the thermal aspects
of the Columbia River system should form an essential part of planning
for ultimate optimum use of the river. Recent contacts by the Bureau of
Reclamation regarding the potential construction of a third powerhouse
at Grand Coulee indicates the desirability of assessing the economic
benefits of thermal regulation, of identifying possible thermal effects
of upstream impoundments, and the need for improving insight into the
physical processes involved. It has been suggested that interested agen-
cies sponsor a cooperative study of the Snake and Columbia River Basins
over the next several years in order to strengthen the theory and pro-
vide a reasonable basis for incorporating appropriate regulating works
in the proposed upstream storage projects.
The General Electric Company, under contract to the Atomic Energy
Commission, currently expects to continue a program related primarily
to operational aspects of the Hanford plant. We hope other agencies
will avail themselves of the opportunity to broaden these studies to
the extent that long-range planning can proceed on a factual basis wirh-
out involving potential danger to resources such as fisheries because
of inadequate investigation.
100
-------
INSTRUMENTATION F'OR WATER-1EMPERATURE STUDIES-1-'
A. M. Moo--e~—
Our knowledge of the thermal properties of lakes and streams has
grown largely with the development of suitable temperature-measuring
instruments and with demands imposed by present and predicted water
use. Temperatures of water in lakes or streams depend upon many fac-
tors, a discussion of which is beyond the scope of this paper. Pro-
fessor Sylvester has mentioned many of these factors in the excellent
paper he presented earlier today.
Instrumentation for t emper'at u-~e determination depends on such
things as the purpose of the investigation, required accuracy of data,
number of additional parameters to be measured, and the depths (pres-
sures) that instruments must withstand. Temperature data are needed
for water-loss, thermal-1oad, water-quality, fish and wildlife, water-
use, and sediment-transport investigations. Equipment presently in
use ranges from the simple, direct-reading hand thermometers to infra-
iled photography. Other agencies may be using instruments with which I
am not completely familiar; my comments will, therefore, be limited
largely to equipment used by the Geological Survey.
Temperature-Profile Recorder
The temperature-profile recorder was developed by the Navy Elec-
tronics Laboratory for use in the water-loss investigations conducted
by the Geological Survey at Lake Hefner, Oklahoma. ]} The temperature-
sensing element for this instrument was a thermocouple which was con-
nected to a length of electrical cable. The small current generated
by the thermocouple was amplified by a storage battery and then routed
to an Esterline-Angus event recorder, which also was used to record
other data. The equipment was maintained in a boat for the temperature
surveys and to conserve electrical power the Esterl ine-Angus recorder-
was clock-driven. The lake thermocouple equipment gave excellent re-
sults once the cables were covered to prevent electrical leakage and
mechanical abrasion* The temperature-profile recorder gave more ac-
curate data than the bathythermograph that had been used previously to
obtain water-temperature data.
"Publication authorized by the Director, U. S. Geological Survey.
'-"-Hydraulic Engineer, U. S. Geological Survey, Portland.
]/ Water-Loss Investigations; Lake Hefner Studies, U.S.G.S.
Profelsional Paper 269, Harbeck and others, 195*+. ~
101
-------
Bathythermograph
In similar water-loss investigations for Lake Mead the bathyther-
mograph was used to record water temperature. The temperature-profile
recorder is more accurate, but would not function properly at the
greater depths of Lake Mead. I/
The bathythermograph provides a continuous record of temperature
versus depth. A stylus, attached to a Bourdon (pressure) tube records
the temperature on a smoked-glass slide. The slide is held in a frame
attached to a pressure bellows and hence the frame and slide move rela-
tive to the arc of the stylus as the depth changes. Thus, pressure
changes resulting from temperature changes cause the temperature stylus
to move, while pressure changes resulting from changes in depth cause
the frame and si ide to move. Observed surface temperatures are used to
calibrate the bathythermograph record.
Whitney Underwater Thermometer
The Whitney underwater thermometer utilizes a small thermister as
the temperature-sensing element and a small dry-cell battery to supply
power needs. The thermister is used as one arm of a Wheatstone bridge
circuit. Depth of observation is measured by the length of line from
the sensing element to the water surface. Temperature is read directly
from the dial of an electrical meter which is calibrated for a range of
5°F. and is provided with multiple settings for temperature range.
The Whitney thermometer was used to obtain water-temperature pro- ,
files in the thermal-load investigations of Lake Colorado City, Texas.-
The instrument, though non-recording, is portable and therefore is
preferable, for some purposes, to the bathythermograph or temperature-
profile recorder.
Infrared Photography
Although the existence of infrared energy has been recognized since
the 17th Century, the development of scientific instruments utilizing
this energy was negligible prior to World War I. Some development of
infrared photography occurred during and shortly after that War, but
significant progress was hampered by lack of sensitive infrared detec-
tors. Highly sensitive detectors were developed during World War II
I/ Water-Loss Investigations; Lake Mead Studies, U.S.G.S.
Professional Paper 298, Harbeck, Kohler, Koberg, and others, 1958.
2/ The Effect of the Addition of Heat from a Powerplant on the
Thermal Structure and Evaporation of Lake Colorado City, Texas, U.S.G.S.
Professional Paper 2728, Harbeck, Koberg, and others, 1959.
102
-------
and these have spurred widespread use of infrared equipment for many
important purposes.
Infrared photography, as applied to bodies of water, utilizes
aerial cameras, infrared detectors, and photographic film. Everything
warmer than absolute zero (-273°C.) radiates infrared energy that can
be optically focused. The infrared scanner consists of a plane mirror
with rotational axis parallel to the line of flight of the airplane.
The mirror scans along a line at right angles to the line of flight of
the airplane. The infrared radiation picked up by the plane mirror is
reflected to a parabolic mirror that then focuses the radiation on an
infrared detector. The detector converts changes in. radiation into an
electrical signal that is amplified and then used to modulate the cur-
rent passing through a glow tube. Light from the glow tube is scanned
across the film in synchronization with the rotating scanner. The film
itself is moved across the exposure station at a speed proportional to
the speed-to-altitude ratio of the airplane. The airplanes are usually
flown at altitudes ranging from 300 to 1,000 ft.
Normally the technique is limited to recording surface tempera-
tures, but the Geological Survey has developed a modification in which
sensing elements placed in the water transmit signals to the airplane
as it passes overhead. In this way, accurate measurements of tempera-
ture of the water mass are made along with measurements of surface
temperature.
Multiparameter Recorder
The multiparameter recorder can record simultaneously many water-
quality parameters including temperature. Other parameters commonly
recorded are dissolved oxygen content (DO), specific conductance, pH,
oxidation-reduction potential (ORP), turbidity, chloride, radioactiv-
ity, and sunlight intensity. In water-quality investigations now being
carried on in the Delaware River and estuary, temperature, DO, pH,
specific conductance, and turbidity are recorded continuously on a
single instrument. I/ Four multiparameter recorders are now being in-
stalled on the Duwamish River in Washington, between Renton and Seattle,
for a Geological Survey investigation of the effect on water quality of
the discharge from a large new sewage treatment plant that is expected
to double the usual low flow of the river below Renton. These four
instruments will record the same five parameters included in the Dela-
ware study and, in addition, solar radiation index (sunlight intensity).
However, not all of the instruments will record all six parameters, but
at two sites some parameters will be measured at more than one depth.
I/ Continuous Recording of Water dual ity in the Delaware Estuary.
McCartney and Bearncr, U.S.G.S., A.W.W.A., October, 1962.
103
-------
The U. S. Public Health Service is using instruments of this type in the
Willamette River, Oregon, to record DO, pH, specific conductance, and
temperature.
In these recorders, a thermocouple develops a small current having
a voltage of 50 millivolts or less that is proportional to the water tem-
perature. The electrical signal is amplified and fed to a strip-chart
recorder or a paper-tape punch. In some instances, the signal is tele-
metered to a central location where receiving equipment provides a peri-
odic print-out of the data and a recording on paper punch-tape.
Telethermpfneter
The telethermometer provides a convenient means of obtaining tem-
perature profiles in streams, lakes, ponds, and reservoirs. As with the
Whitney underwater thermometer, the temperature-sensing element is a
thermister, and a very tiny one, as the probe in which it is housed is
only 3/16-inch both in length and diameter. Power is supplied by two
flashlight batteries housed in a small console. Depth of observation is
measured by the length of line from the thermister to water surface. A
dial on the console provides for registration of temperature directly,
both in degrees Fahrenheit and degrees Celsius (Centigrade). The in-
strument used by the Oregon District does not provide for a recorder,
but some models of the telethermometer do make such provision. Our in-
strument covers a range from 30° to 120°F. with only one dial setting,
but more sensitive (multiple-setting) models are available. We have
modified our telethermometer by constructing a small aluminum bar in
which to house the probe and about one foot of the insulated electrical
cable. This bar provides enough weight to position the probe for depth
in ponded water; the bar can be taped to a wading rod or attached to a
regular hanger bar and sounding weight for use in flowing water.
Thermograph
To obtain continuous records of stream temperature, the Geological
Survey uses a thermograph attachment with the Stevens A-35 water-stage
recorder. The pen trace of water temperature is continuous on the same
strip chart on which is recorded the stage record. Some Federal and
State agencies use recorders that provide a pen trace of temperature on
a circular chart, usually geared to make one rotation in seven or eight:
days. Those recorders are not suited to our normal routine visits to
basic network stations (streamflow and water quality) once every five
to six weeks. Also, the computation of accurate records is more dif-
ficult from the circular charts than from the strip charts. Because of
the difficulties involved in the use of circular charts and the advan-
tage of obtaining both stage and temperature record on the same chart,
we prefer;the thermograph attachment to, the water-stage recorder.
104
-------
The thermograph attachment consists of a temperature probe con-
nected by means of capillary tubing to a bellows mounted on the under-
side of the recorder. The temperature probe is placed at the stream
end of a 1'-z;-inch galvanized inlet pipe and the position of this end of
the pipe is so selected that it is always in moving water. The probe,
capillary tubing, and bellows are filled with methyl alcohol. With
increase in water temperature the alcohol expands and causes a piston-
like movement of the bellows. This movement is transferred to the
temperature pen through a torque arm, gear sector, wheel, shaft, and
beaded cable. The temperature pen, which operates within the top three
inches of the strip chart at a scale of ^0°F. to the inch, is set one
hour ahead of the stage pen to insure that the two pens will never
interfere with each other.
Maximum-Minimum Thermometer
The maximum-minimum thermometer is a U-shaped thermometer so con-
structed that the mercury column on the left side of the U positions a
small metal marker indicating minimum temperature, while maximum tem-
perature is similarly indicated on the right side of the U. The tem-
peratures so designated are those that occur between visits to the
station. A small magnet is used to reset the metal markers on the
mercury columns.
The Geological Survey has completed some preliminary experiments
with these instruments and we believe they show promise as an inexpen-
sive means of obtaining monthly maximum and minimum water temperatures.
Hand Thermometer
Hand thermometers used by the Geological Survey for spot observa-
tions of water temperature are red-liquid-filled and are graduated in
intervals of 1° from 30° to 110°F. Initially, mercury-filled ther-
mometers were used, but the red-liquid type are easier to read and
almost as accurate although slower in responding to temperature change.
These thermometers are also used to check the setting of thermographs.
Hand thermometers that have been accurately tested and calibrated are
used to check the setting of some of the more precise instruments such
as the temperature-profile recorder, bathythermograph, and Whitney
underwater thermometer.
Accuracy of Instruments
Whitney thermometers, temperature-profile recorders and bathyther-
mographs are accurate to within about 0.2°F., but most of these instru-
ments achieve that precision only if they are periodically checked
against a standard milliameter or thermometer.
Thermograph attachments to water-stage recorders are, in them-
selves, accurate only to within about 2°F., but they are checked
105
-------
against hand thermometers and results can be considered accurate to
within about 1°F.
Multiparameter recorders are considered accurate to within 1°F.
The telethermometer used by the Geological Survey in Oregon is
guaranteed accurate within 1% of the range, or 0.9°F. for the 90°
range, but accuracy is generally found to be within about 0.5°F.
Manufacturing practice reportedly permits inaccuracies in hand
thermometers and maximum-minimum thermometers not to exceed one half
of the smallest graduation interval. This means that hand thermom-
eters used by the Geological Survey should be correct within 0.5°F.-
and maximum-minimum thermometers, within 1°F. We have checked our
hand thermometers periodically and errors as great as 1°F. have been
found only rarely.
Precautions in Use and Installation of Instruments
When a water-temperature station is co be establ'i shed, whether it
be recording or non-recording, care must be exercis'ed to see that the
temperature registered is representative for the cross section. For
example, when we installed our first thermographs in Oregon, in 19^9
and 1950, we assumed that if the thermometers or the temperature probes
for thermographs were placed in moving water, the record would be repre-
sentative. A series of near-surface measurements across the section
indicated that this was so. In the summer of 1963, we used a telether-
mometer to obtain temperature profiles at kO thermograph sites. Tem-
perature at the inlet was found within 0.5°f". of average temperature
for the cross section for 3*t of the *40 sites, and within 1.0°F. for 39
of the **0 sites. At one site the difference was 1.6°F., but there the
temperature profile had to be measured about 600 ft. upstream from the
gage. For 35 of the kO sites, the variation in temperature throughout
the cross section was confined to a range of 1.3°, at the other five
sites, temperatures in sluggish water at the banks were found to be
about 2° higher than in the faster-moving water. However, water at
these higher temperatures comprised a negligible part of the flow.
Figure 1 shows the temperature-profile for Breitenbush River above
Canyon Creek near Detroit, Oregon.
A temperature profile is particularly necessary where a tributary
enters a short distance upstream from the proposed water-temperature
station. In such an instance, the waters may not be thoroughly mixed
as they pass the station. A site where incomplete mixing exists is
unsuitable for the collection of water-temperature data.
The temperature probe for the thermograph should not rest on the
stream bed. A few tests made in the summer of 1963 indicated that the
stream bed, when exposed to direct sunlight, can be about 1°F. warmer
106
-------
Water surface
Temperatures are in Degrees Fahrenheit
5
0
rc
(I
CD
30 40
Width in feet
FIGURE I-TEMPERATURE PROFILE FOR BREITENBUSH RIVER
ABOVE CANYON CREEK NEAR DETROIT ON JULY 16, 1963
-------
than the water above it. As mentioned previously, Geological Survey
installations are designed so that the stream end of the temperature
inlet is above the stream bed.
Spot observations of water temperature with a hand thermometer
should be made with the bulb end of the instrument immersed in moving
water as the registration can change rapidly when the bulb is exposed
to air. Because many streams have large diurnal fluctuations in tem-
perature, the time of day should be noted along with the temperature.
Minimum temperatures generally occur about 7 to 9 a.m. and maximum
temperatures about ^ to 6 p.m. Figure 2 shows diurnal fluctuation of
water temperature for selected streams at gaging stations in Oregon.
When maximum-minimum thermometers are serviced, the two readings
should be made quickly because air temperature can reposition one of
the metal markers indicating maximum or minimum water temperature.
Another possibility for error is that the scale for minimum tempera-
ture increases in a downward direction and, so, is easily misread.
Still another chance for error exists when the maximum-minimum ther-
mometer is reset, because the mercury columns and markers must be set
at the existing water temperature. When this has been accomplished,
the thermometer must be quickly placed in position in the water before
air temperature can affect the registration.
Conclus ions
All instruments used for obtaining water temperature can, if used
with proper precautions, yield results that are accurate to within at
least 1°F. Also, all of these instruments have a place in the work to
be done. Temperature-profile recorder, Whitney thermometers, bathy-
thermographs, and infrared photography are adapted to precise measure-
ments needed for water-loss and thermal-loading investigations. Con-
tinuous temperature records should be obtained for routine operational
purposes and where good records are urgently and immediately needed.
Also, at least one thermograph record should be obtained on each major
stream for an indefinite period. Such records could serve as "primary"
records with which "secondary", or intermittent, records could be cor-
related. The secondary records can be from thermographs operated for
just a few years, or maximum-minimum thermometers operated for two or
three summers. Where temperature records are not immediately needed,
the secondary records can be in the form of spot observations made
over a relatively long period. Mult{parameter recorders are desirable
where several water-quality characteristics must be measured. Infra-
red photography is particularly useful in locating sources of pollu-
tion in streams and in locating areas of groundwater inflow. The
technique is also valuable in oceanographic work, where it helps de-
fine ocean currents and circulation patterns-
108
-------
80
70
to
• II
t-
I
< 1
(4
M
0)
60
50
U
( )
a> 80
U
ii
E
* 70
0)
o
60
50
I
r Rogue Rivet near Agness
\DA 3,939
Western Oregon
Stations
South Fork Coquille River
nea r Powers
DA 93.2
Middle Fork Willamette
River near Ou k ridge
DA 258
12PM
12PM
Time of day
6A I2M
12PM
12PM
E aste rn Oregon
Stations
Powder River near
Ric hla nd
DA 1,310
South Fork Walla Walla
River near Milton
DA 63
Metolius R iver
near Grandview
DA 324
DA = Drainage area in square miles
FIGURE 2 - TYPICAL D! URNAL FLUCTUATION OF WATER
TEMPERATURES DURING SUMMER MONTHS
109
-------
DISCUSSION
0_. Do you not use the Ryan thermograph?
A. No. We have written to them and have inquired of some of our
offices who have used them. We are inclined to think that we
probably will try them. They had some trouble over a year ago--
the thermographs were not waterproof—but I think they have cor-
rected most of that now and, if that is so, we are going to try
them.
A. Fourteen of them are used in the Hungry Horse-Libby study with
good results.
A. One of the advantages, of course, is that you don't have to build
a house for these thermographs as we do with others. This is ex-
pensive. They could just be mounted in a pipe and this makes a
very inexpensive installation. That is why we are thinking of try-
ing them.
0_. Tel 1 us again how many of these thermograph stations you have in-
stalled and approximately how long they have been in operation.
A. The first ones were installed in 19^9 in cooperation with the
Oregon State Fish Commission. Only one of these is operating
now. Those installed for the Corps of Engineers have a11 been
in continuous operation since 19^+9 and 1950. Up until this past
summer we had obtained records at about 50 different sites with
33 thermographs which were constantly in operation. This past
summer, in work for the Corps of Engineers, we installed 2^ more.
This was for their 308 Review Report for which these records
were needed. We now have close to 80 thermograph records, 60 of
which are in almost constant operation.
0_. Do you move these around from station to station?
A. Yes. Three or four were discontinued this year in the Alsea
Basin.
110
-------
SUMMARY OF CURRENT THEORIES AND STUDIES
RELATING TO TEMPERATURE PREDICTION
Robert Zeller
Ou t_l j_ne
I. Introduction
A. Paper to be included in the Proceedings* wi 1 1 :
1. Present several methods of temperature prediction
2. Discuss related studies involving these methods
3. Discuss research needs and present a general outline
for a comprehensive reservoir-stream temperature study.
B. This presentation will include:
1. Short description of the usual energy-budget approach
2. Outline of methods involving "heat-exchange coefficients"
a. Computation of "equilibrium temperature" by correlation
with air temperatures.
b. Computation of natural stream temperature input and out-
put functions.
3. A temperature prediction example using an exponential decay
factor.
k. Discussion of research needs relative to the methods of
temperature prediction presented.
II. Energy-Budget Approach
The energy budget attempts to equate the net exchange of heat
between a body of water and its environment to a change in tempera-
ture.
*SA San. Engineer, U. S. Department of Health, Education, and
Welfare, Public Health Service, Water Supply and Pollution Control
Program, Pacific Northwest, Portland.
Ill
-------
The exchange of heat involves: (notation from Schroepfer)
1. The difference between incident and reflected solar radi-
ation (+ ATS)
2. The difference between incident.and reflected atmospheric
radiation and the loss of heat by thermal radiation from
the water surface (i.e., net exchange of long-wave radia-
tion) (- ATR)
3> The loss of heat due to evaporative processes (-&Tg)
k. The gain or loss of heat due to temperature difference at
the air-water interface + (ATC)
5. The heat gain due to discharge, for example, of cooling
water into the reach (+ATA)
These incremental temperatures, then, are added algebraically to
the upstream temperature, TA, to estimate the downstream interface
temperature, Tg, as follows:
-£TR = TB
Other processes actually involved, but usually disregarded, are
biochemical reactions and conduction of heat at the water-channel
bottom interface.
Since computation of evaporation and thermal radiation exchange
depend on the assumed downstream temperature, the equation...cannot be
solved directly. Formulae, such as the above, can be solved by suc-
cessive trials assuming downstream temperatures, Tg.
This method is used equally well on streams in their natural,
steady-state condition as on streams receiving large amounts of
cooling water or cold, reservoir water.
III. Equilibrium Temperatures and Exponential Decay of Temperature Incre-
ments
This second approach is a two-fold operation:
A. First, the steady-state, or equilibrium, temperature of the
water is estimated by any one of several methods.
B. Second, transient temperatures due to thermal additions are
decayed exponentially downstream.
To elaborate just briefly:
112
-------
A. Equilibrium temperatures can be estimated by:
1. Energy-budget equations, as just discussed.
2. Simple correlation with air temperature as Gameson, Hall
& Freddy did on the Thames Estuary in 1957'
e.g.: The Thames Estuary data yielded the following
relationship: Equil. Temp. - & = 0.5 + 1.109 Ta
3. Estimation of the natural stream temperature according
to its response to its thermal environment as expressed
by Outtwei ler ' s equations:
0_n - 25 B,Tad + CBTa x - 4
Heat input - TJt) = ------- + ............ (-----)
f X ^ + CB X
Water temp. = T* (x,t) = T + ^-/ Tunsin (n«/t + rfn-efn)
n - 1
The first of these equations is plotted from a knowledge of cl ima-
tological data. Short time intervals will yield points on a modified
sine curve.
The second equation is merely a reflection of the first modified
by an amplification factor (TU/T.) and phase lag (<0.
B. Regardless of what form the equilibrium temperature takes,
transients can be accounted for by computing the initial temperature
increment and reducing it exponentially downstream. The exponential
decay factor can be expressed as follows: k
— rx
e v
Where: v = average velocity
x = disti downstream
and: k has been evaluated by Major Outtwei ler as A.
z
Where: z = avg. depth
and: \ = C, + C2U2 = 1.35 + 0.2 U2
Where: l^ - estimated wind speed in mph
113
-------
An example of this procedure is worked out in the paper to be
included in the Proceedings.
IV. Finally, I want: to Hsr some of the areas where continuing re-
search is needed relative to the several methods of temperature pre-
diction discussed in the paper included in the symposium Proceedings.
A. Energy-budget approach
1. Inexpensive instrumemat ion to replace the array of equip-
ment currently needed to obtain accurate radiation and evap-
oration data.
One possibility which has been around for some time now is the
CRI. (These are, essentially, insulated evaporation pans instru-
mented to simulate a complete-miniature heat-budget unit. Several
small units of this type were installed recently here in Oregon.
Evaluation results should be available soon.)
2. Extensive correlation o* all available meteorological data
could eliminate the need *or this instrumentation on many
projects, if not most projects.
3> Methods should be developed to determine location and
quantities of significant bank storage.
k. We still need some guidelines on the accuracy of computations
required This relates directly to the basic temperature
criteria problem-
B. Equilibrium temperature correlations and exchange coefficients
1. Although usefulness o* air-water temperature correlation
seems controversial ror m;> regi0r«, a reasonable correlation
study would settle this q^est'o^.
2. The values arrived at by Duttweile' ro^ the constants used ro
estimate the exchange coefficient (^L) need verification.
Synops's
The following paragraphs are intended to familiarize the reader with
some of the currentl y—used methods ro' esrimaring stream temperatures as
a function of their thermal environment and rime No attempt has been
made here to evaluate the methods presented other than to point out major
features of interest.
Ill*
-------
Also included is a presentation of some recent and proposed field
studies involving one or more of the temperature prediction methods*
Several of these studies are discussed in conjunction with the presen-
tation of the method of prediction; others are discussed separately.
Finally, a partial tabulation of research needs relative to the
methods of temperature prediction included in this paper has been at-
tempted.
Summary of Current Theories and Studies
Relating to Stream Temperature Prediction
Generally speaking, there are two currently-used methods of pre-
dicting stream temperatures. The first method applies an energy-budget
approach to both streams in a steady-state thermal environment and to
streams responding to significant discharges of industrial cooling water
or impounded water.
The second method is a two-fold operation. First, the steady-state,
or equilibrium, temperature of the stream is estimated by any one of
several methods, including the energy budget. Second, transient tempera-
tures due to thermal additions are imposed on the equilibrium temperature
profile at the point of discharge and decayed exponentially downstream.
I. The Energy-Budget Approach
The energy budget attempts to equate the net exchange of heat betwee'
a body of water and its environment to changes in water temperature. En-
ergy-exchange processes normally considered include:
1. The difference between incident and reflected solar radiation.
2. The difference between incident and reflected atmospheric radi-
ation and the loss of heat by thermal radiation from the water
surface (i.e., net exchange of long-wave radiation).
3. The loss of heat due to evaporative processes.
k. The gain or loss of heat due to temperature differences at the
air-water interface.
5. The heat gain or loss due to advected water (e.g., heat gain
due to cooling-water discharges).
Other processes actually involved, but usually disregarded, are
biochemical reactions and conduction of heat at the water-channel bottom
interface.
For detailed descriptions of theory and data relative to individual
115
-------
energy-budget parameters, the interested reader is referred to several
references in the bibliography (1, 6, 7» 8, 17).
A. G. J. Schroepfer (1961) presented an energy-budget solution
to temperature prediction for the Mississippi and Minnesota
Rivers at Mi nneapol is-St . Paul, Minnesota (15).
Schroepfer set up the energy budget in the form of its effects
on stream temperature by converting heat-exchange quantities to
incremental temperatures. The resulting mathematical expres-
sion of the energy budget is as follows:
Where TA = temperature of river at point A
^ T/\ = temperature increase due to thermal addition
^ Ts = temperature increase due to solar radiation
Z^ T£ = temperature decrease due to latent heat loss
^ T£ = temperature decrease due to convective heat transfer
^ TR = temperature decrease due to thermal radiation exchange
Tg = temperature of river at point B
After substitution of measured and estimated quantities plus con-
version units, Schroepfer's "working" relationship is:
0.1855 O.OOMt5A
TA +/"-£"-7 HA +/"--Q""7/~Hs - (0-3253) (10 + W) (Vw - Va)
- (0.16) (5 + W) (Tw - Ta) - 1.1 (Tw - Ta)_7 = TB
Where: Q. = river discharge; cfs
HA = thermal additions (i.e., heat load); mega BTU/day
A = water surface area; 1000' s sq. ft.
Hs = solar radiation; BTU/sq. ft./hr.
W = mean wind speed; mph
Vw = saturation water vapor pressure at the mean
temperature of the water surface; mm. of Hg'.
16
-------
Va = partial pressure of water vapor at the temperature
and relative humidity of the surrounding air; mm of Hg.
Tw = mean water temperature; °F.
Ta = mean air temperature; °F.
In the above equation, Meyer's formula for evaporation was selected
since the values for the empirical constants were developed in Minnesota:
E = C (1 + 0.1W) (Vw - Va) = evaporation rate; inches/month
C = empirical constant = 10 for deep rivers to 15 for shallow
streams
W = wind speed; mph
Vw = saturation vapor pressure at the temperature of the water
surface; inches of Hg
Va = absolute vapor pressure at the temperature.and relative
humidity of surrounding airj inches of Hg.
Solar radiation was estimated from St. Cloud, Minnesota, radiation
data after a correlation study using local percent sunshine records for
comparison. No correction for reflected radiation was applied to the
solar radiation data. Wind speeds and other meteorological data were all
estimated from local records.
The above "working11 equation is solved by successive approximations
using assumed downstream temperatures. Ordinarily, convergence on the
assumed temperature is rapid.
With this equation, Schroepfer computed temperature profiles of the
Mississippi River from the Minneapolis waterworks intake downstream 43
miles to the Hastings Oam for the month of August. The results were
checked against Minneapolis-St. Paul Sanitary District temperature data
for the years 1950-59 obtained during routine river sampling operations
(Figure l). Agreement was generally good where the sampling data was
adequate for comparison. Temperature profiles computed.for the Minnesota
River downstream of a large steam generating power plant provided simi-
larly good agreement with measured temperatures (Figure 2).
A six-day, diurnal plot of the Mississippi River temperature a short
distance downstream from a steam generating power plant in St. Paul was
computed for August 25-31, 1959. Measured temperatures at the downstream
station again provided good agreement with the computed profile (Figure
3).
117
-------
00
M
O
tt
en
Crt
C
OQ 13
C >-"
- 3
NO £J
0 TJ
W
H
G
JO
W
en
IQ
C
(P
rt
^
-------
B
en
G
B
££
M- J>
3 £.
3 £
<6 D
o O
*0
2 r
n O
5 g
<-c t1)
•- w
00 Hj
B
t)
O
B
u
cc
o:
LJ
a.
5
LJ
Discharge
Air temperature
Relative humidity
Wind speed
611 cfs
67.7°F
60%
12. 1 tnph
75 Btu/sq
ft-hr
Solar radiation, H
c
-1
to
MINNESOTA RIVER, Miles above mouth
-------
O
O
s
H 1]
W O
01
** »-l
•-» en
CO
o
d
>
r
"Z
O
Sw
- o
!"*
4^ >
H
M
79
12
Noon
Aug. 25
12
Noon
Aug. 26
12
Noon
Aug. 27
12
Noon
Aug. 28
12 12 12
Noon Noon Noon
Aug. 29 Aug. 30 Aug. 31
Measured Temperature Variation
Calculated Temperature Variation
a
re
^>
-------
Finally, the derived energy-budget equation was used to predict
1980 temperature profiles of the Mississippi and Minnesota Rivers based
on estimated steam generation thermal additions and values for meteor-
ological conditions.
B. Jerome M. Raphael (1962) presented an energy-budget formulation
applicable to lakes, reservoirs, and stream increments of isotropic tem-
perature structure (12).
Raphael's method applies a numerical integration of a time rate of
temperature change function as follows:
dtw QtA+mi (ti " tw)
Time rate of temperature change = — = ------------------
d9 nv,
Where: 9 = time
ny, = total mass of the lake
tw = mean temperature of the lake
itif = inflow water mass
tj = inflow water temperature
A = lake surface area
Q.t = total surface heat transfer per unit of time.
Substitution of volumes for mass and simplification for solution over
given increments of time resulted in the following working equation:
lw = 6275 i
Where: V1 = volume of the lake
V^ = inflow volume
tw = average lake temperature over the increment
of time, &
Solution of this equation requires evaluation of the gross heat
transfer parameter, Q.t, as follows:
0-t = <*i - 0-b - °-h - °-e + dv; BTU/sq.ft./hr.
Where: d,- = (1-0.0071 C2) (0_s - Qr) = net short-wave insolation
121
-------
and Qs = incident solar radiation (measured)
Qr = reflected solar radiation (measured or estimated)
C = average cloud cover in tenths of sky covered
Qb = effective back (long-wave) radiation
= 0.9700^(1^ -^T^)
a~ - Stefan-Boltzmann constant =1.71 x 10"'
$ - function of atmospheric vapor pressure
/j
= Qa/0-Ta ; where Q.a « atmospheric radiation
T = water surface temperature; °F.
W
Ta = air temperature; °F.
Qg = evaporation energy = 12U (ew - ea)
U = average wind speed in knots
ew = saturation water vapor pressure at the tempera-
ture of the water surface; in. of Hg.
ea = absolute vapor pressure of water in the air;
in. of Hg.
0_n = conducted heat = 0.00407 UP (ta - tw)
P = atmospheric pressure; in. of Hg.
ta = air temperature; °F.
tw = water surface temperature; °F.
For application to stream temperature prediction, Raphael's working
equation assumes the outflow temperature of a reach to be the inflow
temperature of the adjacent downstream reach.
In his presentation of the method, Raphael includes examples of its
application to temperatures in a small, re-regulating reservoir for a
period of one week and to a major western river, estimating daily temper-
atures for July through September.
C. The 1960-61 Advanced Seminar of the Johns Hopkins University,
Department of Sanitary Engineering and Water Resources, submitted a study
122
-------
report on heat exchange processes in flowing streams (8). This report
includes discussion of the energy budget, turbulent diffusion theory,
and exponential decay of temperature wi th time.
Following a detailed examination of the individual heat exchange
processes, the Advanced Seminar expressed the total "heat budget" for
a selected stream as follows:
S /~QS - Qr - da - d
ar
Where: S = surface area
Q.s = incident solar radiation (measured)
Q.r = reflected solar radiation (measured or estimated)
Qa = atmospheric radiation (measured)
Qar = reflected atmospheric radiation~0.03 Q.a
-------
dt
-- = temperature gradient of air
dz
Y = adiabatic lapse rate
However, using Bowen's ratio, R, Qn = RQe
TO - V P
Where: R = CB £ J
e0 - eg 1000
CB = 0.61 (varies from 0.58 - 0.66)
T0 = water surface temperature; °C.
Ta = air temperature; °C.
eQ = saturation water vapor pressure
at water surface temperature; mb
ea = water vapor pressure of the air;
mb
P = atmospheric pressure; mb
0^ = loss of sensible heat by evaporation of water
at constant temperature
= (f eE) c
-------
Qn = net radiation input
and: reach length = Ax.
average discharge in the reach = q
average velocity in the reach = v = 4*
Then: q(TQ - Tfa) Cp + SQn - SQbs - S( 1 + R) de - SQW +
(dvl1 -Qvo1) = (q - E) (T, - To) Cf
Where: TO = average entering water temperature; °C.
Tb = arbitrary base temperature; °C.
C = specific heat capacity
7 = average water density; gm/cc
T] = average leaving water temperature; °C.
Substituting for: Qbs; Qgj Q^j and (ln + Qe = ( 1 + R) EL/
Then: qC/(TQ - T, ) * EC/(T, -
= S
= S /f(25 + ---) + (1 + R) EL/+ EC/
(T«, - Tb) - as 7
Where: T = To 4- T
This "working equation/1 as with previous energy-budget formula-
tions, is solved by trial and error for the downstream- face tempera-
ture, Tj .
Now, if heat in the form of cooling water is uniformly discharged
into the reach at the upstream face with no significant gain in over-
all discharge, q, then:
- Tb) C/+ qcC/(Tc - Tb) + Sd' n . SQ1^ . S(1 + R1)
ll - a1^!) = (q - Eld (T,1 - Tb) C/
125
-------
Where: qc = cooling-water discharge
TC = temperature of cooling water; °C
Primes indicate new heated condition.
Rearranging and letting: Tj, = To & TQj
Then:
(T,1 - Tb) - E (T, - Tb)_7
+ R)LE 7
(To1 - To) + (T,1-
- c r - ----
L 4
(Tol - To) M
4. <;r r / r1
* bL/ / fc 2
-lV + S/V~1 + RL^1/-
-_/ 1-ay^iTiM.t/
T,1 - To) T, - To
E "" i"" -
By trial and error, this equation will approximate the new down-
stream temperature, assuming a discharge of qc at temperature Tc into
it.
Note that solar radiation quantities are independent of water tem-
perature. Hence, solar radiation data is not needed for solution of the
modified stream temperature. Evaporation data and original temperatures
are needed, however.
In conjunction with their discussion of stream temperature predic-
tion relative to heated discharges, the Advanced Seminar pointed out the
necessity for examining in-stream temperature equalization by temperature
diffusion. For example, if cooling water (or reservoir water) is dis-
charged at a point in a stream, there wi11 be a finite, predictable dis-
tance downstream in which significant vertical and/or horizontal strati-
fication exists. Until a point is reached where no stratification of
temperature exists, temperature equalization by turbulent diffusion must
be investigated because of its effects on heat exchange at the air-water
interface. No attempt will be made in this paper to investigate this
problem further. The interested reader is referred to the Advanced
Seminar and related papers on this subject.
To test their evaluation of the energy-budget processes and study
the relationship of the energy budget to turbulent mixing mechanisms,
the Advanced Seminar analyzed data from an eastern river temperature
study conducted in August, I960.
126
-------
Data collected in the study included 'he following:
1. Location of temperature transect* relative to a heated discharge
outfall as follows:
Transect Djstance from Outfall (ft.)
3 (upstream)
1* 50
6 500
7 1,000
9 3,000
10 5,000
11 7,000
2. Sounding* at each temperature sample point
3. Temperature profiles spaced approximately 100 ft, apart at each
transect. (Stream width averaged greater than 1,000 ft.)
U. Relative humidity along each transect
5. Time intervals of study at each transect (necessary for temper-
ature correction calculations to follow a specific body of water
downstream)
6. Weather description at each transect
7. Automatic stream temperature recorder data in area of the heat
outfall
8. Radiation data from nearest weather station
9. Wind velocity data
10. Heated water discharge and temperature
11. River discharge estimates for the specified date.
Seven transects were studied in all--one upstream of the heat out-
fall and six downstream. Total reach distance was about l1-^ miles.
Field data were analyzed as follows:
1. Advected heat = Q,v - Kfa^^Tjj cal/hr.
Where: a] = subsection areas of uniform width
V] = average velocity at subsection
127
-------
= temperature excess over the natural water
temperature; °C.
K = 0.565 x 108
2. Radiation
The water surface area was divided into areas of equal
temperature where diffusion of the heated discharge was in-
complete. Generally, this involved a warm wedge, natural
temperature wedge, and gradient area.
Net solar radiation input was computed from direct and
diffuse solar radiation given at the weather station,, assuming
the reflected radiation at 0.048 cal./cm2/nrin.
Net atmospheric radiation - 0,97 x 1.25 x solar radiation
Long-wave radiation emitted = 0.97
-------
Biblio.
Ref. Stream
Location
Reservoir
State Author
Study
Sponsoring Completion
Agency Date
10 Rogue R.
Lost Creek Oregon W. Bruce Bureau of
McAlister Sport Fish.
& Wildlife
Clearwater Bruces Eddy Idaho Wayne V.
River Dam Burt
13, 1** Columbia Wells, Wash. J M.
River Rocky Reach, Raphael
Wanapum, and
Priest Rapids
Dams
Walla Walla
District,
Corps of
Engineers
P.U.D. #2,
Grant Co.,
Washington
1961
I960
1961-1962
o-
c
0.
(Q
A
!-*• ™n
rr • i
o
o> —
TO —
= §.
O 3
oi in
-», at
O 1
3 A
in
W
if A
0 <
A
l/> ~l
if O>
A
Q> T
3 A
0
if A
-Q rl"
A 01
n 3
01 CL
if
c n
-I C
A 1
-0 A
T 3
A if
a.
"*• |/)
O if
if C.
-" a.
o —
3 A
M W
-j*
3
to
A
3
A
-i.
3
at
^o
z.
^«
o
01
If
o"
o
in
r*
-s
A
01
3
3
-o
A
-I
a»
^T^
c
T
A
TJ
T
A
O.
O
If
^ft
o
3
•
-------
In addition, the Oregon State Water Resources Board is currently
applying energy-budget analyses to a cooperative study of stream tem-
perature prediction and control in the Umpqua River Basin, Oregon.
The Oregon State Water Resources Board is also applying the energy-
budget method to data collected in August, 1963, on the Coast Fork
Willamette River, Oregon, during a cooperative field study.
II. Equilibrium Temperatures and Exponential Decay of Transient
Temperatures
A. M. LeBosquet, Jr., (19^6) introduced a relationship for deter-
mining heat loss rates from streams receiving cooling-water discharges
(9). The heat loss coefficient, K, would be found from examination of
actual excess-temperature decay rates expressed as follows:
df KASF
dT L
Where: K = heat loss in BTU/sq.ft./hr./°F. of excess temperature
of water over air
F = excess temperature of water over air at distance (0)
milesj °F.
As = surface area; sq. ft.
L = weight of water; 1b.
After substitution of stream physical characteristics, integration,
and simplification, the derived relationship for K is:
Fa
Q. Log,0 F
K = -----
0.0102 WO
Where: Q. = average discharge; cfs
Fa = initial excess temperature of water over air; °F.
W = average stream width; ft.
0 = reach distance; miles
LeBosquet points out in his presentation that the decay of excess
temperature with time is exponential. This phenomenon is readily seen
from a rearrangement of the derived formula for K:
'2.3 (0.0102) WK
= Fa exp< - D
1-30
-------
The range of values for K given by LeBosquet for rapid, shallow
streams and slow, sluggish streams was 18 to 6 BTU/sq. ft./nr./°F.
B. Gameson, Hall, and Preddy (1957) presented a method for pre-
dicting temperature effects on the Thames Estuary due to steam-electric
generation cooling water discharges (5)° Their conclusions were based
on the following study outline:
1. Measurements of water temperatures in the estuary over the
years 19^9-5**.
2. Estimation of "unheated" water temperatures for the same
years.
3. Estimation of the rate of heat entry to the estuary.
k. Calculation of the rate of loss of "excess" temperatures.
5. Prediction of the temperature distribution for unit inputs
of heat at points throughout the estuary.
Elaboration of their study follows:
1. The estuary temperature measurements were taken weekly at a
midstream depth of six feet and at fifteen regularly-spaced
stations over a reach of fifty miles.
2. The "unheated" water temperatures were estimated by regres-
sion analyses of a comparison between long-term records of
water and air temperatures dating back to the first half of
the eighteenth century. This was done for three stations.
The resulting correlations for the three stations studied
were extrapolated to estimate "unheated11 temperature curves
for the entire fifty-mile reach.
3. The estimated rates of heat entry are summarized as follows:
a. Steam-electric power plants (7) 228 x 109 BTU/day
b. Industrial outfalls (gasworks,
paper mills, sugar refineries) 19 x lO?
c. Sewage outfalls 27 x 10?
d. Advected fresh water 20 x 10^
e. Biochemical activity 12 x 10^
TOTAL 306 x 109 BTU/day
b. The temperature change rate for a given heat loss was expressed
as follows:
j0 f
*_ = - f e
dt z
131
-------
Where: 6 = initial temperature increment
f = an exchange coefficient or rate constant with
time"' dimensions
z = average stream depth
t = time
Equating the heat-loss rate to rate of heat entry yielded an aver-
age value for the exchange coefficient, f, as follows:
« d0
&Q. = heat loss rate = -(yz£ x)j0---
dt
and: Q, = J0-f j y€ dx = total rate of loss of excess heat
Where: y = stream width
x = reach increment
j - water density
O" - specific heat of water
The total rate of heat loss, Q., can be equated to the known rate
of heat entry, assuming steady conditions, yielding a value for the aver-
age exchange coefficient, f. For the Thames Estuary, F = k.Q cm/hr. over
a selected period.
5. Exchange coefficients, obtained in the above manner, were then
used in conjunction with a derived temperature discharge distri-
bution method to approximate observed temperature profiles.
C. Gameson, Gibbs, and Barrett (1959) published the results of a
temperature survey and study of the River Lea near London (k).
For a total of 129 hours, water temperatures were measured about
once every hour at five stations below a steam-electric generation plant
and once every three hours at one station upstream.
Air temperatures were measured at each station. Solar radiation
temperatures (mercury-in-glass thermometer with a blackened bulb placed
in an evacuated outer glass bulb) and wind velocities were measured at
a single station.
132
-------
The stream temperature data showed strong diurnal variations of up
to 8°C. near the power plant outfall. Decreasing diurnal variations and
lower mean temperatures downstream indicated definite cooling effects.
Because of the strong diurnal temperature variations, it was pos-
sible to calculate channel volumes and flow times from flow records and
the peak-to-peak time intervals on the temperature profiles.
Equilibrium (or unheated) water temperatures were estimated in the
same fashion as for the Thames Estuary study. Using the same data as
for the Thames Estuary, the equilibrium water temperature was estimated
as follows:
e = 0.5 + 1.109 Ta
Where: £= equilibrium water temperature; °C.
Ta = air temperature; °C.
Again, the exchange coefficient was calculated from the time rate
of temperature change function with the following results:
Over-a 11 average; f = 2.6 cm/hr.
Range of f = 1.66 - 3.83 obtained by averaging all four reaches
over four days
Range of f = 2.10 - 2.88 obtained by averaging all four days for
each reach.
One conclusion of this study, later contradicted by others, was
that the exchange coefficient was apparently unrelated to wind speed.
D. C. J. Velz and J. J. Gannon (1959) presented a comprehensive
paper on stream temperatures; temperature effects on water quality;
magnitude and sources of heat loads; and temperature prediction (16).
Velz and Gannon derived a relationship for the long-term equilib-
rium (unheated) stream temperature from meteorological data as follows:
(1.8 + 0.16W) E + 0.00722 HVC (1 + OJW) VE
= (1.8 + 0.16W) Ta + 0.00722 HVC (1 + 0.1W) Va + Hs
Where: W = wind speed average measured at 25 ft. above the water
surface or surrounding land area; mph
E = equilibrium water temperature (unknown); °F.
133
-------
Hv = latent heat of vaporization at the assumed water
temperature; Biu/lb-
C - constant = ]k for flowing streams of moderate depth
and velocity
VJT = equilibrium temperature vapor pressure (unknown);
in* of Hg0
T'a = average air temperature; °F,
Va - average absolute water vapor pressure of the air at
25 ft. above the water surface; in« of rig,
I-L = solar radiation heat gain (measured); BTU/sq.fto
The above energy-budget relationship is solved by successive ap-
proximation, assuming values for the equilibrium temperature- In the
derivation of this relationship, Meyer's evaporation formula was used
to estimate evaporation.
Velz and Gannon then derived a relaiionship between water temper-
ature and the water surface area required for cooling, as follows:
TZ tfJ
A = -22k, 640 £ ------------- - ---------- - sq.ft. /cfs of streamflow
- VE) +£(TW - E)
Where: C( = 0.00722 HVC (1 t 0,IW)
- 1 08 +• O.I6W
The total increment of temperature between the initial heated
condition and desired downstream temperature is divided into equal
increments (AT'W) with TW and V as the mean temperature and mean satu-
ration vapor pressure in each increment. Note here, however, that E
represents a "long-term" equilibrium temperature and is not specific
for short-time periods- Hence, long- ierm weather data averages are
used in this computation,,
According to Velz and Gannon, the above solution for required
cooling area will yield stream tempera?ure profiles as follows: "Know-
ing the cumulative surface area along '.he course or the stream for rhe
particular runoff from channel cross sectoring, the river temperature
profile for that runoff can be const: "u^ ted J'
E. David W, Outtweiler (19&3) completed a mathematical model of
stream temperature (3)- His derivar'or. began by equating the heat
gained in an incremental reach of *t
-------
n -
-------
After substitution of expressions for the energy-budget terms;
disregarding bank loss; letting wdx = surface area; qe = Ewdx; q =
av (where v = velocity); L = constant; assuming q?^ qe; rearranging
and simplifying, Duttweiler arrived at:
cfa T cfav aT (A + BU2) -
— - + -T-- -- = On - 25 + (Pi Tad + CBTa) if
w ft -~ ^x 2k
( (A
The general expression for evaporation used in the above equation
was as follows:
(A + BU2) (ew - ea) .
E = " > ™d ew - efl =@} (T - Ta(J)
where//] = incremental slope of the vapor-pressure curve.
Note that the above-derived equation is now a space-time expres-
sion for the rate of temperature change in the reach. In this equation,
Duttweiler made the following substitutions:
0 (A + BU )
A = Qn - 25 + 05,Tad + CBTa) —-£---- if', gcal .tfcm2/hr.
„ (A + BU?)
X=1i+ (& + CB) --- if-, gcal./cm2/hr./°C.
' 2^
cPa _T cPav gT
Thus: -<•- ?- + -''-
w dt w
cfa aT cfav &T
and: -J- -- * -i = £ - T
Aw ^t Aw ^x A
Now, let: "z = mean depth = -
w
and: k = -£ =
*.* - 4-
cja cpc
and: Tr (t) = *•
^ A
Then: ?I + v ?I = k/T, (t) - T(x,t)
-}
)C
J
Now, let: v = --
dt
0T a1" dx dT (f ")
Then, by Euler's expansion: -- +---- = — = k/T6 (t) - T(t)i
at 9x dt dt / * )
<— /
136
-------
According to Duttweiler, then, "the time rate of temperature in-
crease is proportional to the deficit between the actual temperature
and some equilibrium temperature." Here, the actual temperature, T(t),
can be described as a reaction, or output, of the stream to the thermal
input, Tf (t). The output temperature, in this case) will approach but
never equal the input temperature.
As seen in earlier discussions, the idea of a proportionality
constant or exchange coefficient, k, is not new. The input function
concept of the equilibrium temperature, however, is believed to be
or i g i na 1 .
Duttweiler solves the above time-rate change of temperature rela-
tionship and arrives at the following expression for the temperature of
a stream as a function of its "equilibrium" temperature:
-kt kt "kt
T(x,t) = e-t etkT (t)dt * e"f ft-
He then represented the equilibrium temperature, T£ (t) by a Fourier
series with a period of 2if hours, wherew= tf as follows:
fni ••—. Mkt
*Y 12
(t) = Tm + n^ , (ancos. nwt + bnsin.nwt)
or a1 ternatively:
op,
(t) = Tm + n4',T1n$in (nut + /$n)
Where: Tm = time averaje temperature
Tin =/'n2 - »2
n
6n = arctan.an
bn
Substituting these expressions for T£ (t) into the equation for the
actual temperature, T(x,t), yielded:
% C- )
T(x,t) = Tm + n r'1 4 an cos. (nu* - qfn) * bn sih. (ntot -
-------
Cfn = arctan. --,
K
or alternatively:
0-5,
"kt
T(x,t) = Tin + , Tun sln.fnwt + ^n - cfn) + e"f(t -2
T1n CH I!
Where: Tun = -- ------- ^ =Ja + b
The input function, T£ (t), is modified by an amplification factor,
Tun/Tin, a phase lag, qfn, and a "transient," fe~'
-------
After substitution of estimated values for the above constants,
Duttweiler determined probable values of C| and C2 as 1.35 and 0.239
respectively. Hence: A = 1-35 + 0.239 U2-
However, comparison with the studies of Gameson, Hall, and Freddy
and of Gameson, Gibbs, and Barrett led Duttweiler to reexamine the
value of C2- The resultsof his reexamination have placed the value of
C2 between 0.179 and 0.239 as a current estimate.
Examination of the previously derived simplification constant, A,
resulted in the following expression:
/«= 0_n - 25 + (Tad + 0.61 Ta) (0.505 + 0.1009 U2)
T C T
= 0_n - 25 +-£4.± --§-?- (y^- l$) = net radiation heat load
+ B
heat load from non-radioactive sources
Duttweiler describesMas "the hourly rate of heating per unit
surface area independent of water temperature."
Then, since T£ (t) = ^:
A ""/?I""CB" " X"
Finally, assuming steady conditions, the water temperature at any
fixed point in the reach, x*, is:
T(x*,t) = Tm + n*r'1 Tun sin. (ntot + jrfn -
-------
- -x
T(x,t) = Tm + Tun s1n.(n0t + /rf -
-------
Estimated Observed Estimated Observed
Month Temp.(°C) Temp.(°C) Month Temp.(°C) Temp.(°C)
Aug. '50 26.92 25.56 Mar. '61 7-59 7.88
Sept. 23.77 22.96 April 12.67 12.16
Oct. 19.66 20.08 May 20.01 19.15
Nov. 11.06 12.37 June 23.82 23.25
Dec. 6.55 6.02 July 26.50 26.77
Jan. '51 2.28 k.37 August .26.92 27-50
Feb. 2.08 5-80
Theoretically, thermograph data would not be needed to estimate
stream temperatures in the situations described above. However, with
current records of climatological data falling considerably short of
the standards required for accurate estimation with the model, thermo-
graph installations are a desirable accessory to the computations.
Initial conditions, at any rate, must be known or estimated.
F. The characteristic of exponential decay with time of tempera-
ture Increments can be put to good use in estimating stream tempera-
tures as a function of regulated releases from storage reservoirs.
The exchange coefficient, as developed by Oavid Duttweiler, was
used to predict temperatures of a reach of the South Umpqua River,
Oregon, as a function of regulated releases from the proposed Tiller
Dam. This example, worked out by Duttweiler, is attached along with
appropriate comments on the method by the writer.
III. Research Needs
The following are a few thoughts regarding research needs of
current temperature prediction theories:
A. Inexpensive instrumentation for obtaining net short-wave and
net long-wave radiation data. The currently accepted method includes
erect and inverted pyrheliometers to measure incident and reflected
solar radiation plus Gier & Dunkle flat plate radiometers to measure
incident, total radiation. Even then, expensive as this instrumenta-
tion is, the thermal radiation from the water surface must be esti-
mated from measurements of the water surface temperature.
The Cunnings Radiation Integrator (6, 17) has been suggested as
a substitute for the pyrheliometers and radiometer; however, the CRI
needs extensive evaluation for the Pacific Northwest region before
its usefulness can be accepted for this area.
\k\
-------
Even with adequate instrumentation, these measurements may have to
be made for individual projects because of the scarcity of applicable
data from nearby weather bureau stations. Extensive correlation of all
available meteorological data at this time may or may not eliminate the
need for complete instrumentation on each project,. Some correlation
studies have already been completed, but much more needs to be done.
B. Extensive water-1oss investigations of the type conducted on
Lakes Hefner and Mead are needed in the Pacific Northwest to evaluate
existing evaporation equations and/or formulate new ones. Again, some
work has already been done by the United States Geological Survey which
is currently involved in mass-transfer studies on McKay Reservoir near
Pendleton, Oregon. However, additional studies are needed at other
regional reservoirs and on the streams themselves.
C. Studies relative to the locatiqn and yield of bank storage
sources both on reservoirs and streams are needed,- This item is usually
ignored, presumably on the basis of its magnitude and temperature dif-
ferences. However, until some method is found to determine these values,
ignoring bank storage may lead to erroneous results,
D. Conduction losses across the water-channel bottom interface
need evaluation. Again, this parameter is usually ignored.
E. The net thermal exchange due to biochemical reactions could
conceivably be of some importance where algal counts are high or pollu-
tion is excessive, but is generally disregarded.
F. A detailed evaluation of the accuracy required of energy-
budget computations is needed. Is it necessary, for example, to pre-
dict temperatures more specific than the mean temperature for the crit-
ical month? How important are diurnal variations in these temperature
predictions? Answers to questions such as these will have a direct
bearing on the complexity of energy-budget instrumentation and compu-
tations. Actually, considerable literature exists on portions of this
problem. An extensive library research could conceivably yield enough
data, if properly organized, to provide reasonable guidelines for
accuracy objectives.
G. The constants used for estimation of the exchange coefficient,
^, need verification (3). This exchange coefficient is a function of
several meteorological variables and is not necessarily constant. The
error involved in assuming a constant value forj^ (and the rate constant,
k = *)needs evaluation.
z
Hf2
-------
H. Simultaneous studies are needed of reservoirs and streams as a
unit. Of necessity, most studies to date have had to be limited to
either a reservoir or reach of a stream. To the writer's knowledge,
there have been no fully-instrumented, long-term studies of temperature
phenomena in a reservoir and its downstream reach as a unit.
Such a study would begin with the selection of a reservoir and
stream physiographical1y and hydrographical1y comparable to other
reservoir-stream units in the same region.,
A complete evaluation of the water budget on the selected reser-
voir stream unit would be essential, TO date, an accurate water budget
is still required to check evaporation estimates and other computations.
Instrumentation should be selected to provide the basis for complete
parameter evaluation. In this respect, complete meteorological stations
are needed both on the reservoir and somewhere along the reach of stream
to be studied. Differences in elevation, ground cover, wind velocities,
and location relative to major geographical relief are all important and
difficult to evaluate when extrapolating meteorological data. If evalu-
ation of prediction methods is the objective of the study, extrapolation
of climatological data from distant weather stations is not acceptable.
There is too much latitude for "fudging the data to fit the answers"
when estimation of radiation or evaporation is involved. In addition to
pyrheliometers, radiometers, wet-dry bulb recorders, anemometers, and
recording rain gages, Cummings Radiation Integrators are recommended at
each meteorological station. If CRI installations are not practicable,
evaporation pans would suffice.
If continuous stage recorders are not already in operation at all
significant points on the selected reservoir-stream unit, they should
be installed where needed.
Thermographs are needed just downstream of the dam, at the crit-
ical reach, and at one or more intermediate points. Additional thermo-
graphs are needed at the reservoir to record significant advected flows
such as major tributaries and at turbine penstocks.
In addition to the full-time instrumentation described above, reser-
voir temperature surveys would be needed at least once every month, pref-
erably once every two weeks. These surveys would include full tempera-
ture profiles down the centerline of the reservoir channel, plus suffi-
cient cross-section temperatures to complete an accurate description of
thermal stratification.
-------
The reservoir temperature profiles would be planimetered and a
mass diagram of heat quantities assimilated to aid in water-budget
computations.
In addition to the routine sampling described above, attempts
should be made to locate and measure velocity currents in the reservoir.
Again, this data would simplify water-budget computations.
Finally, several discharge-depth-velocity surveys should be run
during the course of the study. A fluorescent dye would be used as a
tracer for flow-time determinations.
The total duration of the study would be at least one full year.
It would be preferable, though not necessary, to include two summers
during the study period.
A study following the general pattern outlined above would yield
"working data" for any of the methods described in this paper.
Notes on Stream Temperature Prediction
Assuming a Constant Temperature Source
The following notes pertain specifically to prediction of stream
temperatures at a given point and time based on the exponential-decay-
of-transient-temperatures method developed by Duttweiler. The predic-
tion necessarily assumes that the entire source of flow is fixed at a
short-time constant temperature (i.e., constant for the period of time
under consideration). The temperature of the South Umpqua River, for
example, can be predicted at any time for which the temperature of the
discharge at the proposed Tiller Damsite is estimated. Attached is an
example of the application of this method to the South Umpqua River,
Oregon.
A. Data Collection Requirements
1. Flow data at the damsite, estimated reach of critical water
temperatures (critical point), and above each major tributary. Also,
the major tributaries at their mouths. This data must eventually be
predicted for the period under consideration,
2. Stage-discharge-velocity relationships at the damsite, critical
point, and above each major tributary.*
It would also be desirable to get velocity data at enough inter-
mediate points to assure satisfactory flow-time estimation. Flow times
are critical.
The stage-discharge relationship is essential for parameter esti-
mation in the prediction computations» It would be advisable to take
-------
cross sections and current meter sections concurrently with flow-time,
dye studies.
If possible, at least two surveys, at different flows, should be
made to increase the confidence in the stage-discharge, velocity data
obtained.
3. Thermograph records at the damsite and critical point.
Minimal requirements would be continuous-strip records of water
temperatures for a period of time analogous to the period under consid-
eration. This will usually mean at least several days of records.
It would be better to obtain at least one full year of records to
help understand the general characteristics of the stream relative to
its thermal environment.
Of some value, also, would be graphs at several intermediate
points in the reach.
Note that records must be obtained at the mouth of intermediate,
major tributaries.
U. Wind velocity records at the damsite and critical point for
analogous periods of several years.
This requirement may be difficult to fill in the usual situation.,
The temperature prediction can be completed without this data, using a
conservative estimate for wind velocity.
If anemometers are available in addition to nearby weather sta-
tions, the survey anemometers should be read at both two meters and
eight meters above the water surface. The two-meter data is preferable
for computations. The eight-meter data would be used for correlation
with the weather station data.
B. Data Analysis and Temperature Prediction
1. Basic Methodology
The method developed by Outtweiler assumes that any artificial
deviation from a steady-state temperature will decay exponentially with
distance downstream. In application, this means that with an artifi-
cially-Imposed, uniform-temperature discharge, there will be a variable,
but predictable initial deviation from the hourly temperatures of a
steady-state thermograph at the damsite. These deviations will then
decay on an exponential curve until, at some point downstream, the
stream temperature will again exhibit the natural, steady-state diurnal
variations.
-------
Thus it is, for computation, both the damsite and critical point
thermograph records are needed. The damsite data will yield the alge-
braic value of the deviation at that point. The residual deviation
(after decay) will then be applied to the critical point thermograph
data to yield the predicted temperature curve.
Mathematically, the method is stated as follows:
- ^x
T*(x,t) = T (x,t) + / Ts(t) - T (0,0)_/ e v
(Predicted temp.) = (steady-state temp.) + (temp, deviation) (exponential
factor)
Where: T*(x,t) = predicted temperature at time "t" and mile "x"
taken from the damsite time of "0" and mile "0".
T(x,t) = steady-state temperature at time "t" and mile "x".
Ts(t) = reservoir discharge temperature, noted at Ts(t),
but assumed constant over entire period.
T(0,t --•) = steady-state temperature at damsite. The time
v will be "x" hours less than that used at mile
v
"x" with "v" being the estimated stream velocity
e = Naperian base
k = - (assumed constant for the reach)
WhererX- C| + ^2^2 ~ exchange coeff.; cm./hr.
C1 = 1.35
C2 = 0.18 - 0.2k
IL - wind velocity in mph
i = average stream depth; cm.
v = estimated average stream velocity; mph.
x = miles from damsite to critical point.
2. Application
Note that this method can be applied easily only to the situation
where the entire flow of the stream is discharged from a constant tem-
1U6
-------
perature reservoir during the period under consideration.
Note also that the method assumes a temperature for the reservoir
discharge. If needed, the reservoir temperatures would have to be
evaluated by a separate study. Logically, the solution could be worked
backwards to determine a reservoir temperature-discharge relationship
to satisfy the stream temperature requirements.
For the situation where it is desired to reproduce existing
thermograph records, more sophisticated data collection is required
as follows:
a. Total solar and atmospheric radiation, using a Gier &
Dunkle flat-plate radiometer.
b. Slope of vapor pressure curve within range of water
temperatures concerned.
'£. Relative humidity to obtain dewpoint.
d. Air temperatures.
e. Remaining climatological data is computed or assumed.
This data would be used to estimate, mathematically, the natural
temperature curves. These computed curves would then take the place of
thermograph records.
Temperature Predictions for the South Umpqua River
from Tiller to Winston Assuming Constant Temperature Releases
from the Reservoir at the Proposed Tiller Dam _
Data available for the computations included thermograph records at
Tiller and Winston plus discharge data for March 30 through April 5, 1961,
at several applicable stations and sufficient stream characteristics to
estimate depth of flow and flow times.
From these data, the following quantities were calculated:
°-avg. = 3,600 cfs
Difference in water surface elevations from Tiller to
Winston = 5U1 ft. !
River miles from Tiller to Winston = 55.5
Average channel slope = 1 .85 x 10~3 ft. /ft.
-------
Average channel width = 165 ft.
Average depth of flow = k.9 ft. = 151 cm.
Average velocity = 3 mph (Kutter's n = 0.0*0
The water temperature, as measured by the thermographs, was assumed
to be in a steady-state condition. Hence, the transient, fe~'
-------
To apply the.method, it is necessary to assume a hypothetical tem-
perature for the reservoir release at Tiller. Let Ts(t) = 1+0°F. for
the period under consideration, 00 to 2l+ hours on March 30, 1961.
Finally, some value(s) must be estimated for the exchange coeffi-
cient, X • For expediency, the decay factor, _. k , will be computed
at the same time. e v "*'
> = Cj + C2U2 = 1.35 + C2U2
Where: C2 varies from 0.18 to 0.21+ (assume C2 = 0.20}
U2 = wind speed at 2 meters above the water surface
A= 1.35 + 0.2 U2; also, k = * = -<^-
z 151
Although the usual prediction would involve the critical condition
of low, or negligible, wind speed, a wide range of wind speeds will be
assigned here to show the effect on X- .
U2
(mph)
0
3
5
8
10
15
20
X
cm/hr.
1.35
1.95
2.35
2.95
3.35
^.35
5.35
k
(1/hr.)
0.00891+
0.0129
0.0155
0.0195
0.0222
0.0288
0.0354
t-A
(hr.)
112
77.5
61+. 5
51.3
1+5-0
3U.7
28.2
v/k = 3/k
(mi . )
336
232.5
193.5
153.9
135.0
101+.1
81+. 6
k/v = k/3
(I/mi.)
0.00296
0.00l»30
0.00517
0.00650
0.0071+0
0.00960
0.0118
-t 55-5
e v
0.81+8
0.788
0.751
0.698
0.663
0.587
0.520
Computations for three wind speeds (0, 8, and 20 mph) are shown
on the following pages. These computations and the accompanying graphs
indicate probable temperatures at Winston for the conditions assumed.
To test the hypothesis that estimation of an average velocity, v,
is a critical factor in temperature computations, the estimates will be
recalculated for an average velocity of 2.0 mph.
a = 3,600 cfs
v = 2.0 mph = 2.9^ cfs
w = 165 ft.
A = 3,600/2.9*» .= 1,220 sq. ft.
z = 1,220/165 = l.k ft. = 228 cm.
11+9
-------
Table 1
Temperature Prediction Computations
So. Umpqua River--Ti1ler to Winston
March 30-31, 1961 - v = 3.0 mph
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
Jix d
-------
60
56
o^ 52
H
2
S 48
B
H
H
^
44
4 n
•a g
•~-i
X
h
—•— .
1 1
ESTIMATED TEMPERATURE OF SOUTH UMPQUA RIVER AT WINSTON
For Constant Temperature Of 40°F at Tiller
Equivalent q
V
Travel time
Winston Actual
Winston Estimated (
-~^_
emp
era
>
tur
>
/
e —
/*
—\
/
7
•~- —
^x
- c
A= 2
A= 1
/
-Wi
- H^
^
_nsl
/pol
.35
95
35
~S
:on
:hel
_
Temp.
cm/hr
cm/hr
cm/hr
Acl
:ia
:ua'.
il :
3600 cfs
3 mph
18 hrs
X,
;:•_
- —
^•**»
L T<
fern]
s
s^;
^-s
mp
). i
.\
x
f
/ /
/
/
•
/ ,
/
/
•''
/
/
JV04 08 12 16 20 24 04 08 16
IQ
C
II
-p-
30 MARCH 1961
31 MARCH 1961
-------
U2
(mph)
0
8
20
X
(cm/hr. )
1.35
2.95
5.35
k
(1/hr.)
0.00592
0.0129
0.023**
k/v
(I/mi.)
0.00296
0.0061*5
0.0117
k
- - 55.5
V
e
0.70
0.523
Note at this point that the change in average velocity has had no
effect on the exponential decay factor, _ k re r« An examination of the
~~ y '' * '
units involved indicates this to be logical.
The modified computations for the temperature at Winston are shown
on the following page. These temperatures for Ok to 1600 hrs., March 31,
are plotted on the accompanying graph.
A comparison of Winston temperature predictions for average veloc-
ities of 2.0 and 3*0 mph does, indeed, show considerable variation.
Diurnal variation of the steady-state temperature is important here.
152
-------
Table 2
Temperature Prediction Computations
So. Umpqua River--Ti11er to Winston
March 30-31, 1961 -- v = 2.0 mph
(1) (2)
(3)
(4)
(6)
(7)
(8)
(9)
t T(o,t) T(55.5,t) ff(q,
(Thermograph records)
Oe v
(>-=2.95)
T-(55.5,t) ffe v T*(55-5,t) &e
(3)+(5) 0^=1.35) (3)+(7) 0^=5.35)
(10)
( 55.5,1)
3)+(9)
March 30
00
01
02
03
Oh
05
06
07
08
09
10
11
12
13
43.00
42.25
41.75
41.50
41.25
41.00
40.50
40.25
40.00
40.00
41.25
42.25
43.25
44.75
March 31
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
45.50
45.25
44.75
44.50
44.50
46.00
47.75
49.50
51.00
52,25
53.00
54.00
55.00
55.00
-3.00
-2.25
-1.75
-1.50
-1.25
-1.00
-0.50
-0.25
0
0
-1.25
-2.25
-3.25
-4.75
-2.10
-1.57
-1.22
-1.05
-0.87
-0.70
-0.35
-0.18
0
0
-0.87
-1.57
-2.28
-3.32
43.40
43.68
43.53
43.45
43.63
45.30
47.40
49-32
51.00
52.25
52.13
52.43
52.72
51.68
-2.54
-1.90
-1.49
-1.27
-1.06
-0.85
-0.42
-0.21
0
0
-1.06
-1.91
-2.76
-4.03
42.96
43.35
43.26
43-23
43.44
45.15
47.33
.49.29
51.00
52.25
51.94
52.09
52.24
50.97
-1.57
-1.18
-0.92
-0.79
-0.65
-0.52
-0.26
-0.13
0
0
-0.65
-1.18
-1.70
-2.48
43-93
44.07
43.77
43.71
43.85
45.48
47.49
49.37
51.00
52.25
52.35
52.82
53-30
52.52
153
-------
60
5 6
-.
I
-
c
48
4 4
4 0
IQ
C.
-
-.
3 6
1
,., .
^
^^~^_
, i
1
1
2
/I
/
k
ESTIMATED
For
TEMPERATURE OF SOUTH
Constant Temperature
Equivalent q
V
Fravel time
/
7
_ T
Jin:
= 3600 cfs
^
Z mph
Z8 hr:
i
;ton Estimau
i Her Actua 1
•
pothet
ica
1 1
Temp .
emp
. a
t T
UMPQUA RIVER AT WINSTON
Of 40°F at Tiller
Winston
?d Temp. J
, i
ill
er
)am
sit
•\ctual Tei
i
= 5.
: 2
= 1
35
95
35
np.
cm/hr
cm/
cm/
hr
U *
hr
^-
^ 1
i H
LjL
''/
'
h
//
[
/
/
j
ff
/
^
/
/
04
08
12 16
30 MARCH 1961
20
24
04 08
31 MARCH 1961
12
l€
-------
References
1. Anderson, E. R., Anderson, L. J., and Marciano, J. J., A Review of
Evaporation Theory and Development of Instrumentation, U. S. Navy
Electronics Lab. Rept. 159, February 1, 1950.
2. Burt, W. V., A Forecast of Temperature Conditions in the Clearwater
River Below the Proposed Bruces Eddy Dam, Corps of Engineers, Walla
Walla District, November 30, I960.
3. Outtweiler, 0. W., A Mathematical Model of Stream Temperature,
Dissertation for school of Engineering Science, Johns Hopkins
University, 1963.
k. Gameson, A. L. H., Gibbs, J. W., and Barrett, M. J,, A Preliminary
Temperature Survey of a Heated River; Water and Water Engineering,
63:13+, January, 1959.
5. Gameson, A. L. H., Hall, H., and Freddy, W. S., Effects of Heated
Discharges on the Temperature of the Thames Estuary, Parts I and II;
Combustion, p. 33+, December, I960; p. 37+, January, 1961.
6. Harbeck, G. E., Jr., Kohler, M. A., Koberg, G. E., et al., Water-
Loss Investigations; lake Head Studies. Technical Report, U.S.G.S.
Professional Paper 296, 1958.
7. Harbeck, G. E., Jr., A Practical Fiedd Technique for Measuring
Reservoir Evaporation Utilizing Mass-Transfer Theory, U.S.G.S.
Professional Paper 272-E, 1962.
8« Heat Dissipation in Flowing Streams; Advanced Seminar Report,
Oept. of Sanitary Engineering and Water Resources, The Johns
Hopkins University, June 30, 1962.
9. LeBosquet, M(> Jr., Cooling-Water Benefits from Increased River
Flows, Journal New England Water Works Association, 60:111-6,
June, 19*46.
10. McAHster, B. N., Rogue River Basin Study. Parts I, II, and III;
Water Research Association Report, May 5, 1961: May 15, 1961;
Novanber 22, 1961.
11• Organization for Water Temperature Prediction and Control Study,
Umpqua River Basin. Oregon State Water Resources Board Report,
February, 1963.
12. Raphael, J. M., Prediction of Temperature in Rivers and Reser-
voirs, Power Division Journal, ASCE Proc.; 88:157+, July, 1962.
155
-------
13- Raphael, J. M., The Effect of Wanapum and Priest Rapids Dams on the
Temperature of the Columbia River, Report for PUD No. 2 of Grant Co.,
Washington, September,1961.
]k. Raphael, J. M., The Effect of Wei 1s and Rocky Reach Dams on the
Temperature of the Columbia River, Report for PUD No. 2 of Grant Co.,
Washington,January,1962.
15- Schroepfer, G. J., Susag, R. H., et at., Pollution and Recovery
Characteristics of the Mississippi River, Vol. One, Part Three)
Report by Sanitary Engineering Division, Dept. of Civil Engineering,
University of Minnesota for Minneapolis-St. Paul Sanitary District,
September, 1961.
16. Velz, C. J. and Gannon, J. J., Forecasting Heat Loss in Ponds and
Streams, Journal Water Pol lution Control Federation, 32:392-^+17,
April, I960.
17. Water-Loss Investigations; Lake Hefner Studies, Technical Report,
U.S.G.S. Professional Paper 269, 195*+.
156
-------
ATTENDANCE AT THE TWELFTH SYMPOSIUM
November 7, 1963
Donald F. Amend
Aven M, Andersen
N. H. Anderson
R. L. Angstrom
Robert Averett
Robert J. Ayers
Richard Bakkala
Wil1iam J. Beck
Paul F. Berg
Richard Berg
Harold Berkson
Donald E. Bevan
Russell 0. Blosser
B. R. Bohn
C. E. Bond
Peter B. Boyer
Lt. George Brown
R. L. Brown
Fred J. Burgess
Melvin H. Burke
Roger E. Burrows
Wayne V. Burt
Richard J. Cal1 away
Dale A. Carl son
George G. Chadwick
W. N. Christiansen
Robert F. Clawson
Wm. D. Clothier
A. G. Coche
Chas. W. Coddington
Gerald B. Collin
John F. Conrad
A. C. Cooper
J. P. Cor ley
J. F. Cormack
R. A. Corthell
Frederick K. Cramer
Colbert E. Cushing
G. E. Davis
Wm. H. Delay
Oregon State University
State Shellfish Laboratory
Oregon State University
Oregon Fish Commission
University of Washington
Oregon Fish Commission
Bur. Commercial Fisheries
PHS ShelIfish San. Lab.
Sport Fisheries & Wildlife
Oregon State University
U. S. Public Health Service
University of Washington
Oregon State University
Oregon Fish Commission
Oregon State University
Corps of Engineers
5<+lst M.I. Det.
State Water Resources Bd.
Oregon State University
U.S. Forest Service
Sport Fisheries & Wildlife
Oregon State University
U.S. Public Health Service
University of Washington
Oregon State University
Ore. State Game Commission
Calif. Dept. of Water Resources
Ore. Fish Commission
Oregon State University
Oregon State University
Bur. Commercial Fisheries
Oregon Fish Commission
International Pacific Salmon
Fisheries Commission
General Electric Co.
Crown Ze1lerbach
Ore. State Game Commission
Corps of Engineers
General Electric Co.
Oregon State University
State Water Resources Bd.
Corval 1 is
Brinnon
Corval1is
Portland
Seattle
Portland
Seattle
Gig Harbor
Boise
Corval1i s
Portland
Seattle
Corval1i s
Clackamas
Corval1 is
Portland
Forte Meade, Md.
Salem
Corval1 is
Portland
Longview
Corval1i s
Portland
Seattle
Corval1i s
Salem
Sacramento
Portland
Corval1i s
Corval1 is
Seattle
Clackamas
New Westminster,
B.C., Canada
Richland
Camas
Portland
Walla Walla
Richland
Corval1 is
Salem
157
-------
George R. Ditsworth
Hugh H. Dobson
Peter Doudoroff
Wes Ebel
John E. Edinger
W. E. Eldridge
M. W. Erho
Robert T. Evans
Curtiss M. Everts
El 1iott M. Flaxman
Richard F. Foster
Laurie G. Fowler
John Fryer
Paul Fujihara
Robert L. Garrison
Daniel L. Gerlough
Charles V. Gibbs
J. Wendel1 Gray
A1Ian B. Groves
James B. Haas
James 0. Hal I
J. A. R. Hamilton
George H, Hansen
Gary Hewitt
R. C. Hinchcliffe
Harlan B. Holmes
J. C. Huetter
Jim Hutchi son
Gary W. Isaac
Robert T. Jaske
H. E. Johnson
David C. Joseph
Malcolm Karr
Earl D. Kathman
Max Katz
Kenneth D. Kerri
James T. Krygier
Norman Kujala
R. L. Laird
Robert E. Leaver
U.S. Public Health Service
Oregon State University
Oregon State University
Bur. Commercial Fisheries
The Johns Hopkins University
U.S. Public Health Service
Washington Dept. of Fisheries
State Water Resources Bd.
Pac. N.W. Water Lab., PHS
Soil Conservation Service
General Electric Co.
U.S. Fish & Wildlife Service
Oregon State University
General Electric Co.
Oregon State University
Planning Research Corp.
Municipality of Metro
Pac. N.W. Water Lab., PHS
Bur. Commercial Fisheries
Oregon Fish Commission
Oregon State University
Pacific Power & Light Co.
Wash. Pollution Control Comm.
1870 Fifth N. E.
Gen. Admin. Bldg., Research
Office
Fisheries Consultant
Corps of Engineers
170 S. Owens
Municipality of Metro
General Electric Co.
University of Washington
Calif. Dept. of Fish & Game
State Water Resources Bd.
Oregon State University
University of Washington
Oregon State University
Oregon State University
Oregon State University
A.I.D. India, c/o Dept. of State
Wash. State Dept. of Health
Portland
Corval1 is
Corval Us
Welser
Baltimore
Portland
Vancouver
Salem
Corval1 is
Portland
Rich land
Longview
Corval Us
Rich land
Corval1 is
Los Angeles
Seattle
Corval1 is
Seattle
Portland
Corvallis
Portland
Olympia
Salem
Olympia
Portland
Portland
Salem
Seattle
Rich land
Seattle
Sacramento
Salem
Corval Ms
Seattle
Corval Us
Corval11s
Corvaltis
Washington, D.C.
Seattle
158
-------
Norman Leibrand
Dale A. Long
Harold W. Lorz
W. Bruce McAlister
George McCammon
J. H. McCormick
Arthur B. Mclntyre
Norman J. MacDonald
Barton M. Maclean
Jas. A. Macnab
L. 0. Marriage
Y. Matida
Fred Merryfield
H. W. Merryman
A1 Mil Is
0. T. Montgomery
Phil F. Moon
Albert M. Moore
S. Moriyasu
Kenneth H. Mosbaugh
R. E. Nakatani
Ronald E. Nece
Francis Nelson
Mark L. Nelson
George 0. Nielsen
Anthony J. Novotny
R. T. Oglesby
Waine E. Oien
Melvin J. Ord
Erling J. Ordal
D. L. Overholser
Eben L. Owens
Clarence Pedersen
L. Edward Perry
John C. Petersen
0. C. Phillips
K. S. Pi 1cher
Herbert E. Pintler
Stuart T. Pyle
Edison L. 0_uan
Jerome M. Raphael
Edwin F. Roby
U. S. Geological Survey
Portland State College
Oregon State University
Oregon State University
Calif. Dept. Fish &• Game
Sport Fisheries & Wildlife
U.S. Public Health Service
Corps of Engineers
Corps of Engineers
Portland State College
Soil Conservation Service
Freshwater Fish. Res. Lab.
Oregon State University
State Board of Health
Wash. Pollution Control Comm.
Bur. Sport Fish. & Wildlife
Corps of Engineers
U. S. Geological Survey
Oregon State University
U.S. Public Health Service
General Electric Co.
University of Washington
U.S. Public Health Service
Corps of Engineers
Wash. Dept. of Fisheries
Bur. Commercial Fisheries
University of Washington
Sport Fisheries & Wildlife
Corps of Engineers
University of Washington
Ore. State Fish Commission
Oregon State University
Corps of Engineers
Bur. Commercial Fisheries
Bureau of Reclamation
Oregon State University
Oregon State University
U.S. Public Health Service
Calif. Dept. of Water Resources
State Board of Health
University of California
Sport Fisheries & Wildlife
Portland
Portland
Corval1i s
Corval1 is
Sacramento
Longview
San Francisco
Seattle
Walla Walla
Portland
Portland
Tokyo, Japan
Corval1i s
Eugene
Olympia
Portland
Portland
Portland
Corval1 is
Portland
Rich land
Seattle
Olympia
Portland
Vancouver
Seattle
Seattle
Spokane
Walla Walla
Seattle
Clackamas
Corval1 is
Portland
Portland
Boise
Corvallis
Corval1 is
San Francisco
Sacramento
Portland
Berkeley
Portland
59
-------
Donald L. Ross
Jack Rothacher
Lloyd 0. Rothfus
Roy E. Sams
Roy B. Sanderson
Harold Sawyer
Ralph H. Scott
Robert W. Scabloom
John Seeders
Wm. L. Shapeero
Thomas T. Shen
Dean L. Shumway
R. 0. Sinnhuber
George R. Snyder
Leale E. Streebin
Robert 0. Sylvester
Allan E. Thomas
Edward 6. Thornton
Parker S. Trefethen
William E. Webb
W. Donald Weidlein
E. F. Weiss
E. B. Welch
Henry 0. Wendler
Ray Westenhouse
John W. Wolfe
J. Larry Worley
Boyd Yaden
Franklin R. Young
Stephen A. Young
Robert W. Zeller
U.S. Public Health Service
Oregon State University
Washington Dept. Fisheries
Oregon Fish Commission
U. S. Geological Survey
Ore. State Board of Health
U.S. Public Health Service
University of Washington
Oregon'State University
University of Washington
2309 Wth N. E.
Oregon State University
Oregon State University
Fish Passage Research
Oregon State University
University of Washington
U.S. Fish & Wildlife Service
Oregon State University
Bur. Commercial Fisheries
Idaho Fish & Game Dept.
Calif. Oept. Fish & Game
Oregon Fish Commission
University of Washington
Washington Dept. Fisheries
Weyerhaeuser Co.
Oregon State University
U.S. Public Health Service
State Water Resources Bd.
Oregon State University
U.S. Public Health Service
U.S. Public Health Service
Olympia
Corvail is
Vancouver
Portland
Portland
Portland
Portland
Seattle
Corval 11s
Seattle
Seattle
Corval Us
CorvaUis
Seattle
Corval 11s
Seattle
Longview
Corval1 is
Seattle
Boise
Sacramento
dackamas
Seattle
Vancouver
Springfield
Corval11s
Portland
Salem
Corval Us
Portland
Portland
160
-------