EXECUTIVE SUMMARY
Field Testing and Adaptation of a Methodology
to Measure "In-Stream" Values in the Tongue River,
Northern Great Plains (NGP) Region
by
Ken Bovee
James Gore
Dr. Arnold Silverman
University of Montana
Missoula, Montana
for the
U.S. Environmental Protection Agency
Rocky Mountain-Prairie Region
Office of Energy Activities
Contract No. 68-01-2653
OEA Coordinator: Denis Nelson
Project Officer: Loys Parrish
March, 1977

-------
DISCLAIMER
This report has been reviewed by the Office of Energy Activities,
Rocky Mountain-Prairie Region, U.S. Environmental Protection Agency,
and approved for publication. Approval does not signify that the con-
tents necessarily reflect the views and policies of the U.S. Environ-
mental Protection Agency, nor does mention of trade names or commercial
products constitute-endorsement or recommendation for use.

-------
ABSTRACT
A comprehensive, multi-component in-stream flow methodology was
developed and field tested in the Tongue River in southeastern Montana.
The methodology incorporates a sensitivity for the flow requirements
of a wide variety of in-stream uses, and the flexibility to adjust flows
to accommodate seasonal and sub-seasonal changes in the flow require-
ments for different uses. It also provides the means to accurately and
reliably determine the magnitude of the requirement for each in-stream
use. The methodology is a powerful tool for water managers, providing
the flexibility and accuracy necessary in water use negotiations and
trade-off considerations.
In-stream flow requirements were determined by additive independent
methodologies developed for: 1) fisheries, including spawning, rearing,
and food production; 2) sediment.transport; 3) the mitigation of adverse
impacts of ice; and 4) evapotranspiration losses. In as much as each
flow requirement varied in importance throughout the year, the considera-
tion of a single in-stream use as a basis for a flow recommendation is
inadequate.
The base flow requirement for spawning shovel nose sturgeon was found
to be 13.0 nP/sec. During the same period of the year, the flow required
to initiate the scour of sediment from pools is 18.0 ni^/sec., with increased
scour efficiency occurring at flows between 20.0 and 25.0 m^/sec.
The study shows that an over-winter flow of 2.83 m^/sec. would result
in the loss of approximately 80% of the riffle areas to encroachment by
surface ice. At the base flow for insect production, approximately 60%
i i

-------
of the riffle area is lost to ice. Serious damage to the channel could
be incurred from ice jams'during the spring break-up period. A flow of
12.0 m^/sec. is recommended to alleviate this problem. Extensive ice
jams would be expected at the base rearing and food production levels.
The base rearing flow may be profoundly influenced by the loss of
streamflcw to transpiration. Transpiration losses to riparian vegetation
ranged from 0.78 m^/sec. in April, to 1.54 m^/sec. in July, under drought
conditions. Requirements for irrigation were estimated to range from
5.56 m^/sec. in May to 7.97 m^/sec. in July, under drought conditions.
It was concluded that flow requirements to satisfy monthly water losses
to transpiration must be added to the base fishery flows to provide ade-
quate protection to the resources in the lower reaches of the river.
Evaluation of component in-stream requirements shows that a base flow
of at least 23.6 m /sec. must be reserved during the month of June to
initiate scour of sediment from pools, provide spawning habitat for shov-
el nose sturgeon, and to accommodate water losses from the system. In
comparison, a base flow of 3.85 m^/sec. would be required during early
February to provide fish rearing habitat and insect productivity, and
to prevent excessive loss of food production areas to surface ice forma-
tion. During mid to late February, a flow of 12 m^/sec. would be needed
to facilitiate ice break-up and prevent ice jams from forming. Following
break-up, the base flow would again be 3.85 m^/sec. until the start of
spawning season.
iii

-------
TABLE OF CONTENTS
Page
Introduction 		1
Results and Conclusions		3
Methodological Development and Evaluation		5
Water Quality Component 		9
Fisheries Component 		9
Ice Formation Component 		16
Transpiration Loss Component		18
Sediment Transport Component		22
Research Needs 		24
References		27
i v

-------
LIST OF FIGURES
Figure	Page
1	Component diagram of a comprehensive in-stream flow recommen- 7
dation procedure.
2	Example of integrated methodology used to determine the mini- 8
mum streamflow requirement for the Tongue River.
3	Example of the construction of a* depth contour map.	12
4	Composite Map for the Vial 1 Section.	13
5	Example of preferred area vs. discharge, on "Peak of the	14
curve" plot.
6	Aerial photograph showing canopy cover of the riparian vege- 20
tation.
7	Vegetation map showing canopy cover density as estimated	21
from aerial photo.
8	Movement of fine sand as Bedload in Ft. Keogh Section.	25
LIST OF TABLES
Table
1 Expected ice thicknessess according to current velocities	17
at measurement locations.
v

-------
FIELD TESTING AND ADAPTATION OF A METHODOLOGY
TO MEASURE "IN-STREAM" VALUES IN THE
NORTHERN GREAT PLAINS (NGP) REGION
EXECUTIVE SUMMARY
Introduction
The western United States may be described as a water poor region by
virtue of the fact that the potential rate of evaporation far exceeds the
annual precipitation in large areas of the region. Surface water occurs
primarily as the result of storage of water (in lakes and reservoirs) and
from importation from high precipitation areas, such as mountain ranges.
The importance of surface waters to the development of the West is recog-
nized by the kind of economic activity and population distribution that
sustains the region. Virtually all major population centers are located
near a lake or river.
Because the availability of water is paramount for development, there
may be intense competition for water by different interests. Historically,
water was considered to be used beneficially only if it was diverted for
irrigation, municipal use, or industry. Commonly, applications for water
rights exceeded the total annual flow of the river. It was later recogni-
zed that there also existed certain water demands which required the pre-
sence of a discharge within the natural channel. This discharge was termed
an "in-stream flow", and has only recently been given legal standing as a
beneficial use of water.
Stated simply, an in-stream flow reservation is a water right
held by a state or federal agency, for the beneficial use of fisheries,
maintenance of water quality, maintenance of channel form and process, and
protection of complex riverine-riparian relationships of the ecosystem.
The presence of many simultaneous, and often conflicting water demands,
1

-------
both in- and off-stream, implies the need for a systems approach to water
planning and management. This means not only the identification of the
system to be analyzed, but objectives, boundary conditions, and constraints
associated with each system (Stalnaker and Arnette, 1976). Unfortunately,
the pattern of water planning has often resulted in fragmentation rather
than consolidation of water uses. This has been particularly true in the
case of in-stream flow determinations. While many in-stream flow method-
ologies address the problems of water quality, sediment transport, or ri-
parian vegetation requirements, the final flow recommendation is invariably
based on fishery requirements. Thus, it is implied that satisfaction of
the basic requirements for the fishery will adequately protect all other
in-stream values.
Most of the methodologies used in the assessment of fishery requirements
were developed for cold, high gradient, salmonid fisheries. Experiences
gained from the experimental dewatering of small trout streams may not be
entirely adequate for dealing with a large, complex, warm-water fishery.
Therefore, the validity of extrapolating a salmonid oriented methodology to
a warm water system has been called into doubt.
Furthermore, many of these methodologies are based on the hydraulic,
rather than the biological, characteristics of the stream. In addition, the
methodology may contain assumptions about the biology of certain species, ra-
ther than the demonstrated physical requirements of the species. In some
cases, however, the basis for biological assumptions are rather obscure.
It is further generally assumed that most in-stream uses have an extinc-
tion point; a volume of water below which a given use cannot exist. Method-
ologies which do not incorporate biological responses, or feedback, are fre-
quently unable to define this extinction point. Because there is usually a
2

-------
demand on the planner to optimize the off-channel use of water, the deter-
mination of the extinction point becomes critical. In a negotiation sit-
uation over water rights, the option of a "fall-back position" (for in-
stream flows, the extinction point) is not open. This in turn may either
affect the defensibility and ultimate legal recognition of the flow reser-
vation, or result in preserving a mere vestige of former aquatic resources.
The objectives of this study were designed to solve the types of pro-
blems discussed above. A primary objective was the development and field
testing of a methodology to assess the minimum streamflow requirements for
a warm water fishery. Another primary objective was the methodological de-
velopment and assessment of minimum streamflow requirements of several non-
fishery in-stream uses, and the adequacy with which the fishery flows meet
these needs. A secondary objective involved the development and modification
of data gathering and modeling techniques for large river assessments. Fin-
ally, the study attempts to show how a systems approach to in-stream flow re-
quirements might be implemented and interpreted by a water resource planner.
The study was conducted in the Tongue River basin in southeastern Mon-
tana. Selection of this river was made on the basis of its size, location,
and composition of the fish community. The Tongue River is a medium sized
stream, averaging about 50 meters in width, and draining an area of ap-
2
proximately 18,500 km . The upper river is typified by a fairly steep grad-
ient and cobble bed, which grades downstream to a moderate to low gradient,
and sand-gravel bed, near the mouth. Over 30 species of fish inhabit the
Tongue River during part or all of the year.
Results and Conclusions
The methodology developed and implemented by this study can be a power-
ful management tool, and is a significant improvement over other methods
3

-------
in current use. The methodology incorporates a sensitivity for the flow
requirements of a wide variety of in-stream uses, and the flexibility to
adjust flows to accommodate seasonal and sub-seasonal changes in flow re-
quirements for different uses. It also provides the means to accurately
and reliably determine the magnitude of the requirement for each in-stream
use.
We found that different in-stream uses assume positions of greater
or lesser importance during different times of the year. Flow requirements
for fisheries are of highest importance (i.e. require the greatest amount
of streamflow) only during certain portions of the year.
Winter ice is one condition which periodically requires a sufficient
flow to prevent serious problems. The thickness of the surface ice sheet
was found to be inversely related to the velocity of the water beneath it.
Therefore, reduced streamflow would not only reduce the depth of water, but
actually give rise to a greater thickness of ice. At the minimum fish rear-
ing flow, up to 90% of the riffle areas of the stream could be solidly fro-
zen from surface to bed-. At the minimum flow requirement for maintenance of
insect productivity, only about half of each riffle area would be so frozen.
During the spring break-up of the ice sheet, serious damage from ice
jams may occur unless the streamflow is fairly high. These jams are vir-
tually assured if the flow is at the minimum level defined for fisheries.
It is estimated that a flow of approximately three times the fishery flow
is needed to protect the stream and irrigation structures or equipment from
damage by ice jams along the Tongue River.
Pool scour during the spring runoff period is considered essential to
maintain stream habitat areas, and to remove silt from diversion structures
and irrigation pump intake areas. The flow required to initiate the removal
of coarse sediment from pools was found to be considerably higher than the
4

-------
flow required for spawning during the same time period. It was further
determined that the efficiency of water use for this purpose could be im-
proved by providing a flow above the base level for pool scour. Extreme
care must be taken during this time period to avoid deposition of coarse
sediment in areas where it doesn't normally occur, for this can lead to
streambed armoring, a condition that makes the initiation of scour more
difficult in succeeding years.
Perhaps the most striking discovery of this research was the deter-
mination that water losses must be added to the base in-stream flow to
ensure that flow at all points of the river. We found that the instantan-
eous flow requirements to satisfy evapotranspiration for riparian vegetation
and irrigated crop land often exceeded the rearing flow for fisheries by
200 to 300 percent. If water is lost from the system, it is absolutely
necessary to account for it. Failure to do so will result in large reaches
of river with less discharge than required for other in-stream uses. In
fact, approximately 50 km. of the lower Tongue River would have zero dis-
charge if only a fishery rearing flow were released from the Tongue River
Dam during July and August. However, this in-stream flow requirement ap-
plies primarily to losing rivers. Where a river is augmented by groundwater
accretion, the problem is not so serious, for the groundwater will also sup-
ply a large part of the water for evapotranspiration. It is apparent that
most in-stream flow methodologies were developed for gaining rivers, as no
methodology in current use even mentions or considers the problem of tran-
spiration losses.
Methodological Development and Evaluation
The "In-Stream Flow Methodology" developed and implemented in this study,
should be conceptualized as the end result of the application of several pro-
5

-------
blem specific, component methodologies. The relationships between the com-
ponent methodologies and the final flow recommendations are illustrated in
Figure 1.
Some of the in-stream uses described in Figure 1 are considered comple-
mentary; that is, the streamflow requirement to satisfy one use will also
satisfy several others. Under this framework, the complementary use with
the highest flow requirement for a certain time period is considered critical.
Consumptive water uses, such as transpiration losses from natural ripar-
ian vegetation or irrigated cropland, are considered additive uses. Flow
requirements determined for these water losses are summed, and added to the
critical flow requirement as determined for the complementary uses. The to-
tal in-stream flow requirement for a given time period is then defined as
the summation of the additive and critical complementary requirements.
The use of the comprehensive methodology to make a minimum streamflow
recommendation is illustrated by the flow diagram in Figure 2. For the month
of June, the flow requirement for pool scour was determined to be 18 m^/sec.,
3
while the base flow for spawning shovelnose sturgeon is 13m /sec. The flow
required to flush fine sediment from gravel areas is only 11 m /sec. Thus,
the critical streamflow requirement for all complementary uses is 18 m^/sec.
3
Total transpiration losses were calculated to be 5.58 m /sec. Therefore,
3
the total in-stream commitment for June is 23.6 m /sec. This is the instan-
taneous flow to be released at the Tongue River Dam to ensure a flow of 18
3
m /sec. at the mouth of the river. The total monthly volume commitment would
be 63.2 million cubic meters.
This methodology is based on the determination of the minimum streamflow
requirement to satisfy each in-stream component; therefore, it is extremely
important that each of these determinations is accurate and valid. Each
6

-------
Figure 1: Component diagram of a comprehensive in-stream flow recommendation procedure.
Components at the far left of flow diagram refer to minimum flow requirements
for each component. For some months, certain components go to zero.
Monthly Riparian
Vegetation Require-
ment
Monthly Irrigation
Requirement	
SuSDended Load
Bedload
Early Spawning
Species Requirement
Intermediate
Spawning Requirement
Late Spawning
Recui rement
Food Production
Cover
Habitat Area
TDS
Total
Monthly
Transpiration
Loss
Greatest
Sed. Trans.
Requi rement
Greatest
Monthly
Spawning
Reaui ren.ent
Greatest
Monthly
Rearing
Requi rement
Dissolved Oxycien

Greatest


Monthly
Temoerature

Mater Qal.


Requi rement
Monthly
Add iti ve
Recui rement
Largest of
Complimentary
Requi rements
for each
Month
Ice
Formation

Greatest
Ice
Break-up

Monthly

Ice ['litigation



Requi rement
Total
Monthly
Instantaneous
] F1ow Requi rement
Total
Monthly
Volume
Commi tment
ii mmum
Monthly
F1 ow
Hydrograph

Total
Annual
Volume
Ccmmi tmeni

-------
Figure 2:
Example of integrated methodology used to determine the minimum streamflow
requirement for the Tongue River during the month of June.
CO
Monthly
Instantaneous
F1 ow
Requi rement
23.6 cms..
x t i me
(2.6784 x 10 sec/month) (June)
To Id 1 monthly volume
commi tment
6 3
63.21 x 10 m /month
U
o
CO
CO
D
CO
LT>
Sediment transport (18.0 cms.
Sturgeon Spawning (13.0 cms.)
Hater Quality
(11.0 cms.)
Iranspiration losses
(5.58 cms .)
3
1	Pool scour requirement
2	Entrainment disncarge of fine sediment
3	Corrected for rainfall
* cms. = cubic meters per second
Riparian (0.9 ens. )
Irrigation (4.68 cr.s .

-------
methodology used in this study was evaluated by the criteria of accuracy,
ground truth or biological feedback, and water availability. A brief des-
cription of each of the component methodologies, and an evaluation of their
validity, are provided below.
Water Quality Component:
The methodologies employed in the determination of flow requirements
to maintain water quality may be found in general sanitary engineering texts,
such as Velz (1970), or in technical bulletins outlining specific water qual-
ity problems. The water quality component was not implemented or evaluated
in this study due to constraints of manpower, and the general lack of pollu-
tion in the Tongue River. However, evaluations of the different water qual-
ity models are usually given with the model descriptions.
The importance of different water quality parameters will vary from
stream to stream, depending on the waste loading, geology, land use, and
water use. Therefore, it is important to identify the most critical water
quality parameters for a particular stream, and the time period during which
that parameter is likely to become a controlling factor in the determination
of streamflow requirements.
Fisheries Component:
The methodology used to determine minimum streamflow for fisheries is
designated the "Critical Area-Indicator Species Method". This method util-
izes the high sensitivity of shallow water riffles and gravel bars to changes
in streamflow. Because of the rapid change in hydrologic variables with
change on discharge, these habitat areas reflect a greater degree of change
than other stream habitats.
A logical extension of the critical area concept is that organisms
which inhabit or utilize these shallow water areas will necessarily be more
9

-------
sensitive to changes in streamflow than will inhabitants of less sensitive
habitat areas. Therefore, if suitable habitat is maintained in the criti-
cal area, for species using that area, it is assumed that conditions will
be suitable for all other species in the river. This concept is used to
define the criteria for an indicator species.
Under this methodology, the same indicator species is not used through-
out the year. Rather, several indicator species have been identified for
use during the migration and spawning phase of the life history, and others
during the rearing phase. For the purposes of this study, the sauger and
shovelnose sturgeon were selected as migration and spawning indicators.
Rearing conditions were evaluated in terms of productivity (aquatic insect
habitat) and balance (habitat diversity for fish). Maximum biomass and di-
ver i sty were utilized as criteria for maximum productivity. The stonecat
was used as the indicator species delineating critical area habitats.
The fisheries component method involves two separate phases of field
data collection. The first phase is the formulation of flow criteria for
each indicator species. This flow criteria includes a determination of
the velocity range, depth range, and bottom type most commonly associated
with each indicator species.
The second phase evaluates a series of discharges across a critical
area, or areas. This evaluation process requires the construction of plan-
imetric hydrologic contour maps which delineate areas of equal depth or vel-
ocity. Field data for the maps collected from four or more transects across
a critical area. At specified intervals across each transect, measurements
of depth, velocity, and bottom type are made. A depth contour map is pre-
pared by transferring the transect depth measurements to their appropriate
positions on a scale drawing of the channel. Lines of equal depth are then
10

-------
drawn on the map by interpolating to the desired depth contour between
points on the same transect, or between corresponding points on adjacent
transects. Figure 3 is an example of a depth contour map with the field
data left on the transect lines. Velocity maps are similarly constructed
on a separate planimetric map. These maps may be read in a manner similar
to a topographic map.
The adequacy with which a given dishcarge provides habitat over the
critical area for the indicator species may then be evaluated with the use
of a composite map. Composite maps are constructed by superimposing the
appropriate depth, velocity, and bottom type contours on the same map.
Only the contours delineating the boundary conditions for the flow criteria
of a given indicator species are presented. Areas which do not meet depth,
velocity, and bottom type criteria are then cross-hatched. Areas remaining
on the composite map meet all of the flow criteria for the indicator species.
These areas may then be measured with a planimeter, and the total area meet-
ing the flow criteria determined for each discharge. The area meeting this
criteria is then plotted against discharge for each discharge "mapped" (Col-
lings, et. al., 1972). This curve is used to determine the optimum and mini-
muni streamflow requirements for the indicator species. An example of a com-
posite map is given in Figure 4. Figure 5 shows a plot of preferred habitat
area vs. discharge.
According to the widely used procedures, the minimum streamflow require-
ment is defined as 75% of the optimum, as determined by the peak of the curve.
However, we found that this was not a'valid approach. The optimum flow has
no direct bearing on the minimum flow requirement, and this procedure often
results in a flow recommendation which does not meet the criterion of water
availability. For example, the minimum rearing flow requirement as determined
11

-------
1	1
5 Meters
Figure 3 : Example of the construction of a depth contour map. Handwritten numerals are from
field data; isolines of equal depth drawn by interpolation between data points;
depths are in cm.

-------
Figure 4 ; Example of a composite map, showing areas
meeting flow criteria.for the stonecat.
Area with no cross hatching meets all
cri teri a.
Total area meeting criteria: 425 m^.
Discharge = 12.0 cms.
5 Meters
Area not meeting depth criteria
Area not meeting velocity criteria

-------
2 3 4 5 6 7 8 9 10	20 30
Discharge in cubic meters per second
Example of preferred area vs. discharge, or "peak of the
curve" plot used to determine optimum and minimum stream-
flow requirements.
14

-------
by the 75% method was found to be equalled or exceeded less than 40% of
the time from 1961 to 1970 for the Tongue River.
Therefore, a new approach was developed to objectively identify the
minimum, or base flow, requirement. The base flow is defined as that dis-
charge which first provides some increment of preferred hydrologic condi-
tions (i.e. meets the flow criteria of the indicator species) over the
critical area. The base flow, thus defined, was equalled 'Or exceeded
more than 70% of the time between 1961 and 1970 for the Tongue River.
The field tests conducted- to determine biological feedback for the
methodology gave credence to both the concepts of optimum and base flow.
As the base flow is approached, it would be expected that diversity and
productivity might decline. In fact, these were the findings of the field
tests. The change in diversity and productivity may be quite precipitous
as the flow falls below base flow level. In addition, we noted some subtle
qualitative changes in the composition of the fish community, which had
serious implications for recreational fishing. When the flow decreased be-
low the base flow level, suckers and minnows became dominant, with a corres-
ponding decline in the number of sport fish present.
Based on the findings of the field tests, we conclude that the base flow
concept, as used within the context of this methodology, was accurate, valid
and met the criterion of water availability. However, the utility of this
method is likely to be limited by the large time and manpower investment
needed for implementation. Another potential objection to the methodology is
that it is only applicable to small, wadeable rivers.
Both objections to this methodology may be effectively overcome through
the use of specialized equipment and hydrologic flow prediction models.
Two such models were field tested during this study. The Water Surface Pro-
15

-------
file Program (WSPP) is an energy balance model, utilizing the Manning equa-
tion to predict depths and average velocities of a channel cross section
at different flows. The CONTOUR model, developed during this study, util-
izes power functions based on hydraulic geometry to predict depths and
water column velocities at specific points on the cross section. While
neither model was able to predict the exact flow parameters, both show con-
siderable promise as flow predicting tools. The CONTOUR model was designed
especially for use in planimetric mapping, and was found superior to WSPP
for velocity prediction within the constraints of the model. Output from
CONTOUR may be used in conjunction with computer plotting programs, so it
may be possible, in the near future, to use computerized simulation mapping
as a technique for in-stream flow work. It is estimated that such simulation
mapping will reduce the time requirement for the use of the Critical Area-
Indicator Species Method by as much as 80 percent. This technique would be
applicable to any river, although boat mounted stream gaging equipment would
be required on larger streams.
Ice Formation Component:
At the outset of the study, it was recognized that ice formation played
an important role in Northern Great Plains streams, but the extent and nature
of its influence was not known. We found that ice formation actually
affects the river ecosystem in two ways: 1) by reducing the amount of habi-
table area by its physical presence, and 2) by disturbance of the bed during
the break-up period.
Field measurements showed that the equilibrium ice thickness (the thick-
ness at which heat loss from the top of the ice sheet equals the heat gained
at the water-ice interface) was inversely related to water velocity. Thus,
slower water velocities were invariably associated with a thicker surface
16

-------
ice sheet. As velocity decreases drastically with reduced discharge, it
follows that a decrease in discharge would result in a greater equilibrium
ice thickness.
The extent of habitat loss to the formation and thickening of an ice
sheet depends not only on the discharge, but on the thermal regime of the
region as well. The methodology used in this study to assess the area
loss to ice utilizes an empirically derived relationship between ice thick-
ness and current velocity. For a given water velocity, there is a corres-
ponding expected ice thickness. Examples of this relationship are given
in Table 1.
Table 1: Expected ice thicknesses according to current velocities
at measurement locations.
Current velocity	Equilibrium thickness
cm/sec	cm.
60	19
50	21
45	21
40	22
30	25
20	29
In as much as the data from Tab'le 1 was empirically derived, it is applicable
only to the Tongue River, although the velocity-thickness relationship may
hold for all rivers which experience surface ice conditions.
The extent of habitat loss may be evaluated for the same critical areas
used in rearing flow determinations, by using the same depth and velocity
contour maps prepared for those areas. Velocity contour lines are used to
estimate the approximate thickness of the ice. This expected thickness is
then superimposed on the depth contour map for the same discharge. Where
17

-------
the predicted ice thickness is greater than, or equal to the depth within
a given contour interval, that area is assumed frozen to the bed. The
approximate percentage of the riffle lost to ice formation may then be es-
timated, or actually measured if greater accuracy is desired.
The determination of the flow requirement to prevent ice jams from
forming is considerably easier than determining the fishery flow. Since
the greatest thickness of ice measured empirically was not much larger
than 30 cm., it was assumed that most of the ice released from pool areas
would be around 30 cm. thick. The depth contour maps for a given critical
area were examined to determine the extent of the area less than 30 cm. in
depth at various discharges. If the depth over the entire area is greater
than the thickness of the ice floating through the area, the chances of
ice jam formations are decreased considerably. Conversely, the probability
of a serious ice jam over a critical area increases as more of the critical
area is less than the required depth.
At this time, the relationship between ice thickness and current velocity
can best be determined empirically. However, there are several thermodynamic
models for the prediction of equilibrium ice thickness which may be of value
in the near future (Faily, et. a 1., 1 975; 1 976). These models were developed
primarily for the determination of ice cover in streams receiving thermal
effluent. They may also be applicable to unheated streams if the heat pro-
duction and abrasion of running water can be determined. These models are
now unverified for in-stream flow use.
Transpiration Loss Component
Since transpiration losses along a river are additive, it is not possible
to select a critical area for the determination of a satisfactory flow require-
ment. The methodology used to determine transpiration losses is essentially
18

-------
the same for riparian vegetation and irrigated cropland.
The first step, in either case, is the determination of the total can-
opy cover involved in the transpiration process. For irrigated crops, this
area is usually available in the literature (i.e. U.S.G.S. streamflow rec-
ords and gaging station descriptions). However, the area of riparian vege-
tation may be determined empirically through the use of aerial photographs.
For this study, vegetated areas were traced from aerial photos onto a grid-
ded tracing paper, in such a manner that each small square of the grid con-
tained some canopy. The percentage of each square covered by canopy was es-
timated, and that vegetation area was given a weighting factor depending on
the percentage cover. The total cover for each vegetation map so prepared'
was determined by measuring each vegetated area with a planimeter, and mul-
tiplying by the appropriate weighting factor. Total canopy cover was deter-
mined by summation of individual canopy cover areas as determined for each
map. The construction of a weighted area vegetation map is shown in Figures
6 and 7.
The transpiration rate was then determined using the Penman equation,
an energy balance equation, and local climatological data. Using the Penman
equation, it is possible to determine the mean daily and maximum daily trans-
piration rates of the vegetation. These rates may be corrected for rainfall
if desired.
Daily water volume requirements for transpiration may then be determined
by multiplying the total canopy cover area (in square meters) by the trans-
piration rate (meters per day). Instantaneous flow requirements were estima-
ted by assuming that groundwater recharge is constant over time. By dividing
the mean daily water volume requirement by 86,400 sec/day, it is possible to
calculate the instantaneous mean daily flow requirement for transpiration.
19

-------
r* 45
I >'•*&'-
/
h ^
!i-.: £***>-
Figure 6: Aerial photograph showing canopy cover of the riparian
vegetation of the Tongue River floodplain, in the vicinity of Ash Creek.
Photo courtesy of the Montana Department of Fish and Game.
20

-------
Figure 7: Vegetation map showing canopy cover density as estimated
from the aerial photo (Figure 6) of the Tongue River floodplain near
Ash Creek.
21

-------
Predicted instantaneous flow requirements were tested against ground
truth data as determined by stream gaging records for the Tongue River.
It was determined that the Penman equation, in conjunction with weighted-
area vegetation maps, was accurate to within +_ 10% using data available
for the Tongue River. With a sufficient number of meteorological stations,
or with site specific meteorological data, it is likely that the accuracy
could be improved considerably. Given the procedural problems of other
empirical methods for determining instantaneous flow requirements, this
method is at least as accurate and time efficient as the others.
Sediment Transport Component:
The primary objective of the sediment transport component is to ensure
that sediment accumulation in pools is removed by a scouring flow. In this
case the pools are considered critical areas. Additionally, the implemen-
tation of this methodology requires the determination of a critical stream
reach, which is defined as the reach with the greatest sediment source, and
the least ability to remove sediment. This reach may vary from stream to
stream depending on the nature of the sediment inflow. For many streams,
however, the critical reach for sediment transport will be in the lowest
part of the basin, where the hydraulic gradient of the stream is at a mini-
mum.
Sediment transport was evaluated empirically by taking both bedload
and suspended sediment samples at the downstream end of a large pool in the
critical reach for the Tongue River. Sediment rating curves for various
size fractions of sediment were constructed by plotting the sediment load
against the discharge. The interpretation of these sediment rating curves
is often made more difficult by a scattering of the data points for a given
discharge. This scattering phenomenon is most serious for particle sizes
22

-------
less than 62 microns in diameter. However, this size fraction may be omit-
ted from the evaluation process because it does not reflect channel scour
or flushing. Rather, these size particles are already in transport when
they reach the main stem of the river. Fine to medium sand, either moving
in suspension or as bedload is better indication of the initiation of scour.
Theoretically, scour may be indicated in two ways. If the load of sed-
iment leaving the channel section is greater than the amount entering the
section, then scour may be assumed to be in progress provided it can be es-
tablished that sediment is not entering the channel within the section. An
alternative method is the examination of the competence of the stream. This
involves the determination of the discharge at which certain size fractions
begin moving. The mass balance method works well if all sediment sources
entering the critical stream section can be measured. However, if bank
caving or slumping of material is a large source of sediment for the stream
section, the source is difficult to quantify. Competence is equally difficult
to determine, as areas within the channel may move some sediment particles at
virtually all discharges.
The method used to evaluate sediment transport in this study is a combin-
ation of both the competence and mass balance methods. Numerous samples were
taken at a relatively low flow, with no sediment entering the river from run-
off. These samples were used to establish "background" levels of the load of
the various size fractions of sediment. Subsequent samples at higher dischar-
ges, on the rising limb of the hydrograph, were then examined to determine
whether the load appreciably exceeded the background level. Thus, a type of
cluster analysis was used to interpret the initiation of scour in the pool.
For example, the background level for fine sand moving as bedload was found
3
to be around 3 metric tons per day for a number of flows below 18 m /sec.
23

-------
3
The load at 18 m /sec. varied from 4 to 5 metric tons per day. However, from
18 to 20 m3/sec. the fine sand load increased to 10 to 12 metric tons per day.
This relationship is shown in Figure 8. It was therefore concluded that pool
scour was initiated at about 18 m3/sec. This flow was also substantiated by
larger particle sizes moving as bedload, and by the appearance of fine sand
in the suspended load.
3
It can be seen from Figure 8, that the value of 18 m /sec. is extremely
conservative for a recommended pool scouring flow. Since pool scour is only
initiated at this flow, it is suggested that a slightly higher discharge,
3
perhaps 25 m /sec., would give a more efficient use of water for this pur-
3
pose. Under the defined objectives of the study, 18 m /sec. is considered
the base flow, and would be used as a fall back position in a negotiation
situation.
There are numerous hydraulic models which could be implemented to de-
termine the flow requirements needed for sediment transport. The simplest
of these is the competence, or critical tractive force, model. This model
derives the discharge at which a certain size of bed material begins to
move. However, it can be seen from Figure 8 that some bed material is vir-
tually always in motion, without the occurance of scour. Therefore, the
competence model is considered to be of little value for in-stream flow
work. Other sediment transport models incorporate a mass-balance aspect
which could be of considerable value. These models require much the same
type of field data as the Water Surface Profile Program, a factor which
adds to the attractiveness of the mass-balance model.
Research Needs
In the foreseeable future it may be possible to simulate any of the com-
ponents of Figure 1. The development and refinement of predictive tools
which reflect real-world conditions accurately, should be given high priority
24

-------
1000
500
200
100
50
20
10
5
2
1
8 :
2	5	10	20
DISCHARGE IN CUBIC METERS PER SECOND
50
1C
Dvement of Fine Sand (125 to 250 Microns) as Bedload in the
t. Keogh Section, Tongue River, Montana
25

-------
for future research. The use of computers for data simulation and for sim-
ulation mapping can reduce the time required to implement a large scale
study from a year to two, to from one to three months. The benefits of
such an approach in terms of cost and number of studies that can be comple-
ted in a limited time frame, is enormous.
Considerably more information regarding flow criteria for different
species is needed. The techniques involved in the implementation of the
Critical Area-Indicator Species approach are applicable in virtually every
river. However, the important species in different regions, or in different
sizes of streams, may change. There is no universal indicator species for
in-stream flows. Therefore, it is important that indicator, or target, spe-
cies for different regions be identified, and flow criteria for those spe-
cies developed.
Further field testing of methodologies should continue on a long term
basis. Important aspects of such studies should include the degree of deple-
tion of a fishery under an incremental flow reduction below the base flow le-
vel, long term effects on the species composition and age class strength, for
flows near or below the base level, long term effects on sediment transport
and channel morphology, and long term effects on the zonation of riparian ve-
getation .
The mechanisms of surface ice formation and its relationship to discharge
are only partially understood. Research into the prediction of equilibrium
ice thickness and the thermodynamics of running water should be continued. An-
other unknown in terms of surface ice that merits study is the effect on stream
organisms.
26

-------
References
1.	Collings, M.R., Smith, R.W. and Higgins, G.T., 1972, "The hydrology
of four streams in western Washington as related to several Pacific
salmon species", U.S.G.S. Water Sup. Pap. #1968. 109 pp.
2.	Paily, P.P., Kennedy, J.F., and Dagan, G., 1975, "Effects of Large
Thermal Discharges on Ice-Covered Rivers", la. St. Wat. Res. Inst.
Ames, Iowa. ISWRRI-68. 10 pp.
3.		., and Macagno, E.O., 1976, "Numerical prediction of ther-
mal regime of rivers", J. Hyd. Pi v., ASCE, HY3: 255-274.
4.	Stalnaker, C.B. and Arnette, J.L., 1976, "Methodologies for the deter-
mination of stream resource flow requirements: an assessment", USFWS,
OBS, 199 pp.
5.	Velz, C.J., 1970, Applied Stream Sanitation. Wiley Interscience, New
York, 619 pp.
27

-------