\ .
’ ba~ «WM ”-0..-

DEIERmmATI-GN OF NECEESARY SPILLWAY CAPACITY

WYLIE A. BOWMASTER

mam

Iv”—

 

 

DETERIIELTIOH 0F NECESSARY SPILLIAI CAPICIT!

by 1-

.nj\

Vylie Li Bow-eater

1 133818
Submitted to the Greduate School of lichigan
State College of Agriculture end Applied

Science in pertial fulfilment of the
require-eats for the degree of

CIVIL ENGINEER

Department of Civil Engineering
1940

THESIS

-——-\_

TO
IR. 3. B. [IIBALL

himipel Hydraulic Engineer
with the Tenneseee Valley Authority
when technical inetructien and guidance
during the past six years have helped to
mete at proper appreciation of the
reepeneibility ofthe engineer,
thie theeie ie reepeeti'nny dedicated

. r} pun ..

max) I i)

m

Intredneticn

Purpo.. e e e e e e e e e e e e e e e e e e e e

”M1. 0 O O I O O O O O C I O O O O O O

Cal-onllethode
Introduction . . . . . . . . . . .

I'p1r1¢.1 Bountione e e e e e e e e e e e e e
'QXiIu- 11°°d Of aflcord e e e e e e e e e e e
Frequency of Pest Floode at the Site . . . .
“hit Greph e e e e e e e e e e e e e e e

lexinue Recorded Ratee of Runoff over Sieiler
B‘tianll Rulfifr‘u.th0d e e e‘e e e e e e e e

Applicetion in.!enneeeee River Basin
Deter-inetion of lexiene Flood
teeneeeee River at Chettenooge
Introduction e e e e e e e e e e e
DOBcrlptlon of ‘h‘ 38813 e e e e e
Aveileble Hydrologic Dete . . . . .
Application of Empirical Equetione
Iexinnl Obeerved Rates of Runoff .
Frequency of Pest Floods .. . . . .
n‘tionfll RuDOff ‘.th0d e e e e e e
"Ii-u. Flnod EIdrongph e e e e e

Hieeeeee River at Hireeeee Dee Site
Introduction e e e e e e e e e e e
DUCCriptian 0f the Basin e'e e e e
‘ylilabl. Hydr01ogic Data e e e e e
Applicetion of Empirical Eqnetione
Frequency of Pest Floode . . . .
Unit Graph UBthOd e e e e e e e e
lazilnn Observed Betee of Runoff
RationAI RUDOff a‘thOd e e e e e
”331'“. F1°°d Hydrogrlph e e e e

0....

Deter-inetion of Deeign Flood
GCHDTCI e e e e e e e e e e e e e e e e
Hineeee River at Hieaeeee Den . . . . .
Generel Obeervetione . . . . . . . . . . . . . . .
selected Bibliogrephy . . . . . . . . . . . . . .

AGEBOIIOngIatI C e e e e e e e e e e e e e e e e

e o e e e e e e e ale e e e e

14
16

25
28

57
59

45
49
59

65
65
65
66
67
68
69
71
76

78
85
B7
90
91

.Qassziaiiaa
Topographic flap, Upper Tennessee River Basin

Iean Annual Rainfall, Upper Tennessee River Basin

Tennessee River Gages at end near Chattanooga, Tennessee
Distribution of Floods, Tennessee River at Chattanooga, Tenn.
Distribution of’aydrologie Stations, Upper Tenn. River Basin

laxinue Observed Flood Discharge Rates, Selected Areas in
Eastern United States

lazieue Runoff Rates, Eastern United States
pylood Frequencies, Tennessee River at Chattanooga, Tennessee

Frequency Curve, Tennessee River at Chattanooga, Tenn.
Free Hydro—Electric Handbook, Creager a Justin

Frequency Curves, Tennessee River at Chattanooga, Tennessee
Free Eater Supply Paper 771

Frequency Curves, Tennessee River at Chattanooga, Tennessee
Free Ester Supply Paper 771

Frequency Curves, Tennessee River at Chattanooga, Tennessee
Pros Hater Supply Paper 771

Probable Flood Frequency Curves, Johnsonville, Florence,
Chattanooga, Knoxville-H.S. 528/7l/2

Relation betseen SOD-year Flood and Drainage Area
a.n. szs/7i/2

Rational Iethod, Tine Zones, Upper Tennessee River Basin
Ieteorology, Tennessee River Basin above Chattanooga

Typical Stores, Cloudburst or Thunderstorn Type, Tennessee
Valley'Region

Distribution of Rainfall, Western North carolina Sotrn of
July 14-16, 1916

West Indian Hurricane Los Pressure lovenents and the
Corresponding Rainfall Areas

Outstanding Stores, Occurrence and Paths of Great Rainfall

Exhibit
Nunbe;

NH

10

11

15

14
15
16

17

19

2O

§§§I§12§ {Continued}

tion

Cyclonic Stores, Low Pressure Hovenents and the
Corresponding Rainfall Areas

3:30.817. Storl Rainfall, 1889, 1895, 1897, 1906, and 1910
EXOOCI1VI store filinfsll, 1915, 1916, 1917, 192?, and 1929
EIDOIIlV. Store Blinfall, 1956 and 1957

Tine—Ares~Depth Curves, Eastern United States

Rsinfhll and Runoff Graphs, Principal Floods at Chattanooga

Transposed Position, Upper Tennessee River Basin, Store of
larch 22-27, 1915

Estineted Bydrograph, Tennessee River at Chattanooga, Tenn.
Transposed 1915 Store

Trenspoeed Position, Upper Tennessee River Basin, Store of
January 12-25, 1937

Estisated Hydrograph, Tennessee hiver at Chattanooga, Tenn.
Transposed 1957 Stors

lees-Duration Curves, Historical and Hypothetical Floods,
Tennessee ‘iver at Chattanooga, Tennessee

Bydrogreph, laxisus issused Flood, Tennessee River at
Chattanooga, Tennessee

lazies- Recorded Runoff, Hieessee—Ocoee River Basins
Distribution of floods, Hisassee River near Reliance, Tenn.

Transposed Position, Store of larch 22-23, 1929, Bieassee
River*Baein

lasisun lssuned Flood, Hieassee River, Hieessee Dan, 8.0.
Ares, Capacity & Discharge Curves, Hinessee Project
Routing Diagram, Hivassee Des, 5-hour Intervals

Exhibi

Number

21
22
23
24
25
26

31

S2

53

55

INTRODUCTION

m

later has often been referred to as Ian's greatest friend and
his vorst enemy. Asple support of the first claim is suggested by the
location of every great city on a navigable waterway, vhile the cold
truth expressed by the latter half of this trite phrase has been for-
cibly deeonstrated in recent years by the disastrous floods vhich seept
northeastern United States in 1955 and 1956, and by the great floods on
the Ohio and lississippi Rivers in 1957. In the present covenant to
conserve the natural resources of the United States, such disasters
have given impetus to investigations for the conservation and control
of the water resources and have resulted in the construction of a
nunber of storage dams and the planning of a great nany more. These
projects range in size free the small earthern structures built by
farmers to check erosion on their fields to the huge multiple-purpose
projects and systems constructed by various governnental agencies to
provide integrated development and control of the actor resources of
an entire drainage basin. A knowledge of the naxinun flow which each
of these structures say be called upon to pass during entrees rainfall
and runoff conditions is vital in assuring the safety of the structure
itself and the life of the whole area which it is designed to benefit.

It is the purpose of this paper to present and discuss briefly
the sore cannon sethods shich have been used to detersine the spillsay
capacity that should be provided in such eater control projects and to
illustrate then by application to points in the Tennessee River basin.
The author has assisted in sisilar investigations made in the Water

Control Planning Depart-eat of the Tennessee Valley Authority, and

funds-ental data and nethods available with the Authority have been
utilised in the preparation of this thesis. Conclusions drawn and
opinions eXpressed, however, are those of the author and not of the
Tennessee Valley Authority, although they may be in substantial agree—
lent in some cases.
The {goblga
' The design of any structure intended for impounding flowing
water lust include an energency spillway for releasing flow exceeding
that which the structure is designed to store. If the failure of such
a structure were to result in only a limited amount of property damage
and no loss of life, the spillway eight be designed to carry a flow
which is likely to be exceeded only once in 25 or 50 years, as it
would be more economical to rebuild the structure and pay damages than
to provide the expensive construction necessary to make it safe against
any flood which night occur. lost of the structures with which we are
concerned, however, are not of this type-they are planned to benefit a
large area and a large number of peOple, and fialure would result in
great prOperty damage and possible loss of life. Thus, they must be
designed with spillways sufficiently large to carry with safety any
flood which eight reasonably be expected. A large spillway is expensive,
eSpecislly when the limitations of the site are such that it is diffi-
cult to find space sufficient for spillway, navigation lock, and power-
house, and the designing engineer is faced with the problem of providing
the necessary spillway capacity and yet keeping the design practical
and the costs within reason.

If the spillway provided in a water control structure is

subjected to a larger flood than it was designed to carry, the headwater

I
elevation must rise to an elevation above that considered in designing

the structure. This is not necessarily serious in a cancrete structure,
since the design usually provides ample factors \f safety which will
cover such a contingency, but if the velocities produced below the dam
are sufficiently high to cause excessive erosion at the foot of the
Spillsay, the structure may be weakened to the p~int of failure. In the
case of an earthern structure, headwater appreciably above that considered
in the design, although below top of the structure, may weaken the
structure and cause failure. Failure is sure to occur if the headwater
rises sufficiently to overtop any portion of the earth banks, as a
slight flow over the top produces sufficient erosion to permit an in-
creased flow nhich then increases the erosion, and so on until the
entire structure fails.

Failure of a small water storage or power dam during the
passage of a flood is apt to cause some damage to roads and buildings
immediately below it, but its failure may be pertioulrrly serious if it
creates a sizesble flood wave which may travel downstream to another dam
and cause it to be overtoPped, and so start e chain of failures which
grows progressively worse as the wave proceeds downstream.

The destruction resulting from the failure of a eater control
structure generally increases greatly with the size of the structure and
the amount of water which it impounds, and due to the large volume of
water stored and suddenly released, is particularly disastrous in the
case of a structure built to control floods. The loss of life may be
enormous, especially if a flood control reservoir of considerable storage
capacity is built to protect a city which then continues to expend over

the river's flood plain formerly considered unsafe for anything but those

4
industries which could stand an occasional flooding, but now considered
secure in the protection afforded by the upstream reservoir. The
property damage in such a situation would also be enormous, and it would
have been better if the structure designed to control floods and to
protect the city had never been built at all. Damages to highways and
railroads paralleling the river would be for greater for some distance
dcvnstream than the damage which might have been produced by the uncon-
trolled flood itself, and the interruption to business caused by the
destruction of the transportation system iould be a very real loss, even
though it might be difficult to estimate. Although the loss of the dam
itself and probably the powerhouse is not to be overlooked, there is also
the loss of the use of these structures over the time required for their
rebuilding. lore important is the loss of the confidence of the people
of the surrounding area and possibly throughout the whole country in
the organisation operating the structure, this lack of confidence
probably affecting the success of that organization over several genera-
tions in the future.

‘ Another less serious but very important result of a hesdwater
elevation above that considered in the design is the damage to private
preperty along the river above the dam. The height of a water control
structure is often fixed to a large eitent by thexpresence of buildings,
roads, and other improvements along the edges of the proposed reservoir
which would have to be bought or relocated should a higher den and pool
elevation be adapted. A flood which produces a heedwater elevation '
sufficient to damage such preperties may result in damage suits and
litigation which are expensive in dollars but particularly expensive in

the loss of the good will of those concerned.

5

Spillwsy capacity can be provided only by increasing sub-
stantially the total cost of the project, and in some cases the cost of
providing a spillwsy of preper sise nay be so great as to make the whole
project entirely unJustifisble. It is desirable that the spillway dis-
charge directly into the natural channel of the river, although in many
cases this is not possible and a side channel nust be provided either
through the ebutnents and thence back into the river or as open channels
around the sbutsents. Small floss say be discharged through gate-controlled
openings in the base of the dam itself, but this method is too expensive
to be considered in dealing with the discharges that must be handled
during a flood.

If power is to be develOped at the project, the channel portion
of the river lust also provide race for a powerhouse. Discharge from
the turbines must be carried away as rapidly as possible in order to
keep teilwster elevations low end to provide the noxious possible head
for power production, and direct discharge into the original river
channel is the most satisfactory way of doing this. In the case of a
navigable stress, e sizeable portion of the channel is occupied by a
navigation lock to lift boats over the den. Since sufficient depth must
be svsilable below the lock’for boats to enter even during periods of
low flow, the lock should be located in the deeper portion of the channel
and east be aligned with the channel below to permit boats to enter
with ease. Thus, the natural river channel must provide room for a
powerhouse and s navigation lock, and since the densite was probably
chosen because of the narrow valley and river channel at this point,
the width remaining for the spillwey is apt to be 83511. To obtain

greater spillway width, either the powerhouse or the lock must be moved

6

out of the channel. This is likely to result in both on expensive and
en inprecticsl arrangement. The poeerhouse might be moved downstream,
but a {lone or tunnel would be necessary to convey ester free the reser-
voir to the turbines, and unless additional head eere to be gained by
this arrange-ant, the cost would be excessive. The navigation lock night
be located out of the old river channel, but this would require the
reeorsl of enoreous quantities of earth and would be likely to result
in poor align-out of the lock and the navigation channel below. Other
spilleay arrangements might be made, such as tunnels through the shut.
eents, Open channels around the shutsents, sluices and siphon spillwaye
through the den, etc., but usually these are more expensive than an
open spillesyt

Iith the spillwey length controlled by the natural width of
the river channel and the specs that must be elloted to the powerhouse
and navigation lock, increased dischsrge capacity may be obtained at a
given heedester eleVetion only by lowering the elevation of the spilleey
crest and supplying crest gates or by increasing the height of the gates
originally considered. The height of gates is limited by considerations
in their design. The 50-foot gates proposed by the Tennessee Valley
Authority for Kentucky Dan on the Tennessee River are among the highest
in the eorld. As gate heights are increased, the intensity of pressure
at the bottoe of the gate increases as well as the total pressure on the
gate. The greater gate height results in greater eeight, while the in-
cressed pressures near the bottom require a stronger design which further
adds to the weight. Increased weight of the gate and increased pressure

of the gate on its supports both lend to more powerful lifting mechanisms

7
and more coeplicated designs to reduce friction, while the increased
pressures else require more elaborate gate seals. The increased thrust
of the gate against its supporting piers requires that the piers be
stronger, while the pressure differential on the two sides of a pier
created by an ogen gate on one side and a closed gate on the other
side result in another series of forces-all tending to increase the
width of the piers and thereby reduce the net length of the spillway.
Home of these difficulties are eliminated by using roller or drum gates
in place of the usual Tainter or Stoney gates, but these may not be
practical for great depth and their costs soon become excessive.

It is apparent, therefore, that the engineer designing a
water control project is often faced with the difficult problem of
providing e epillway which will carry with safety any flow to which it
say be subjected, at the same time keeping his design practical and his
costs within reason, realizing all the time the disaster that night
result should he underestieate the flood-producing ability of the drainage
basin or allow hisjudgnent to be influenced by requests for a cheaper

design.

ggyhON hETHODS
Introduction

A consideration of the problem of determining the spillway
capacity that should be provided at a water control project indicates
that it can be divided into four parts:

(1) Determination of a discharge hydrcgraph representing the
laxilul flood reasonably to be expected at the site under the conditions
existing prior to construction of the project or any other water control
project within the drainage area above. This may be termed the “maximum
expected flood, natural conditionsI or just the “maxi-um flood.‘

(2) Determination of a discharge hydrograph representing the
greatest flood reasonably to be expected at the site under the conditions
existing just prior to construction of the project. The effect of up.
stream regulation is here considered. This represents the greatest
-flood which must be considered in planning the project and may be termed
the “project flood.“ The project flood nay bathe maximum expected flood
as reduced by storage in upstream projects, or it may be the result of
a very heavy store centering over the uncontrolled drainage area
illediately above the project.

(5) Determination of the maximum outflow and the maximum head-
water elevation that are likely to result under the proposed scheme of
operation of the project or under any operation that is likely to take
place. This may he termed the "design flooi,‘ inasmuch as it represents
the eaxieue flow for ehich the project must be designed. It will result
fro: routing the project flood through the reservoir, consideration
being given to the effect of the natural and controlled storages within

the reservoir upon the project flood. It is here that the necessity

of considering a project flood rather than Just a project discharge
because apparent, as the maxi-us outflow and maximum hesdvster elevation
ere dependentrupon,the total volu-e of flood flow as sell as the maxi-us
rate of flee.

(4) Detersinstion of the spillvey dimensions which will most
closely fit the functional design of the project. The spillwsy of a
flood control project shich does not have sufficient storage capacity
to retein the entire flov of s single greet flood not only nust pass
the design flood st s high or maxi-u- heedvster elevation, but it also
lust pass rather large floss ehen the pool is at a low elevation in
order that the pool say he quickly drsvn down upon the approach of a
greet flood to provide additional flood storage. It should else be
possible to hold the pool st s cooperatively loe elevation during the
pesssge of the first portion of s flood save, thereby reserving con-
siderable controlled storage space for use near the crest of the flood
ehen this storage can be used to the greatest advantage in reducing
the flood crest. This investigation involves routing floods of verious
types through the reservoir to note the effect of various heights of
gates end lengths o spillvsy upon the effectiveness of the flood control
operetions of the reservoir._ Speed with vhich the reservoir say be
emptied after it has been filled by e flood nay enter the problem, since
it say be necessary sometimes to quickly release the stored flood esters
in order to provide storage space for s second flood.

The design flood will have s noxious discharge considerably
less than that of the project flood if the controlled storage is lsrge,
end there say be some reduction if there is no controlled storage, but

othe uncontrolled reservoir storage increment is greeter than the natural

10
channel storage increnent for a given increase in discharge. The tern
'chennel storage increment' as here used represents the values of eater
stored temporarily in the natural channel of the river vithin the limits
of the proposed reservoir when a certain change in flow takes place under
given conditions of flee. “Uncontrolled reservoir storage increnent' or
feet "uncontrolled storage increment" represents the volune of water
stored teeporarily vithin the reservoir when a certain change in flow
takes place under given conditions of flov with all Spillsay gates open.
The uncontrolled storage increment is normally greater than the channel
storage increment since the reservoir area is generally much greater
then the natural stress area, while the spillway usually limits the £10!
to e certain extent and forces the reservoir surface to rise almost as
rapidly as it sould rise under natural river conditions to accossodate
a certain increase in flow. ”Controlled storage' represents the voluee
of eater stored under given conditions of flow in excess of the uncon-
trolled reservoir storage under the same conditions of flow. This
values may be released or stored at sill by opening or closing the spill-
esy gates, whereas the uncontrolled storage cannot be regulated by gate
Operation.

The design flood may have a naxinun discharge equal to or even
slightly greater than that of the project flood if there is no controlled
storage under project flood conditions and the uncontrolled storage
increment is not appreciably greater than the natural channel storage
increment. This simplifies the problem to a determination of the project
flood, and since flood volume does not enter the study, the problea is
further simplified to the anxious flow of the project flood. Since nest
of the projects built for power alone belong in this class, the practice

in the past often has been to compute the maximum flood to be expected

11
at the site, considering any reduction in flow produced by the reservoir
or upstream reservoirs as a factor of safety but neglecting the possi-
bility of discharges being increased by the reservoir. Accordingly, a
large number of equations and practices have been develOped for
estimating the maximum expected flow or the flow for which structures
should be designed. In the case of flood control reservoirs and multiple
purpose reservoirs as constructed in recent years, the large amount of
space reserved for retarding; or storing flood eaters may result in a
design flood substantially smaller than the project flood or the maximum
eXpected flood, and more elaborate studies must be made to arrive at
the most reasonable outflow for vhich a spillvay must be designed.

Like most other hydraulic investigations, there is no simple
nethod which can be applied to all projects to determine the naximun
expected flood or the project flood. The judgment of the designing
engineer must always detersine the final answer-but there are many
lethods which he may employ to assist his judgment. The difficulty is
tounderstand the advantages and limitations of each method in applying
it to a specific project. Certain of the methods cannot be employed
because of lack of necesssry data, but lack of time should never be an
excuse for neglecting any of the methods. There is too much at stake to
omit any study which might shed additional light on the problem.
Empirical Equations '

The tern "empirical equations" is here used to cover the
multitude of formulas expressing maximum flood flow as a function of
size, or size and shape, of the drainage basin. The most commonly used

type considers drainage area alone, and may be written as Q 3 C An,

there 'A' is usually the drainage area in square miles, ”n' tn eXponent

12
varying fro: ebout 0.5 to 1.0, and "C“ a coefficient which may very fron
about 5000 to 10,000, depending upon the various factors influencing
flood runoff, such as type of stern, season of the year, tapogrephy, shape
of the basin, vegetation, valley storage, and channel storage.

Several equations of a slightly different type were formulated
by luichling from A study of drainage areas up to 5000 square siles on
the loheek River. Kuichling gives the equations Q =~%§;4§98. {'7.4
for rare floods, and Q 2 “Singgg. / 20 for occasional floods.

A similar countion is prepcsed by lurphy for drainage areas
up to 10,000 square miles from a study of Her Englenl streams:

Q :.§§:Z§%b / 15.

Those eguetions which attenpt to include the effect of factors
other than drainage area are generally lengthy and too involved to com—
pare without 8 detailed plotting. The equation of C. R. Pettis, pub-
lished privately in 1927, gives Q : c r I1-25 in which to” in flood
discharge in cubic feet per second, "F” is the average width of the
drainage area, 'P' is a rainfall coefficient, and '0' is a coefficient
representing the combined influence of ell the factors not mentioned in
the other terns. The rainfall coefficient is the loo-year pluvial index
of e 6—day store as computed in the studies of the Miami Conservancy
District for each of the quedrengles into which the eastern half of the
United States was divided. The basin width is the average width obtained
by dividing drainage area by the length of the main portion of the river.

A similar equation was proposed in 1929 by F. G. Switzer and
»H. 0. Killer, with an adjustment in the rainfall factor to allow for
size of drainage area and differences in the time of concentration.

0f the various equations involving frequency, that published

by I. E. Fuller in 1914 is the best known. There are really two

15
equations. The first states c; = o", (1 ,l 0.8 loglo r) in which ~03".
is the average sexisue annual flood, and ”Q” is the probable maxi-us
annual rate of flow during the period of years 'T.' The average maximum
annual flood is to be obtained from a record of stages at the site, or
if this is not available, from the second equation proposed by Fuller,
which states Qa'. = c 10‘8 in which 'A' is the drainage area in square
eiles and '0' is a coefficient whose value depends upon the flood pro-
ducing characteristics of the drainage area. The«xnation thus reduces
to o .. c 10-8 (1 ,l 0.8 loglo T). It should be noted, however, that the
discharge given by this equation is the probable maxi-u- ennuel flood
to be expected in a given number of years andnot the flood to be
equalled or exceeded in the given number of years. The latter conception
is included in sost equations involving frequency and in nest of the
frequency methods to be discussed later.

Another type of equation that should be sectioned here is that
in which intensity of rainfall is one of the factors considered. The
egaations is generally written Q 8-0 i A, in which 'A' is the drainage
area in acres, "i‘I is the intensity of rainfall in inches per hour,

'0' is a coefficient of runoff, and 90' is the discharge in cubic feet
per second. This equation is often spoaen of as the 'retional eethod'

end is used in computing the runoff for which city stern sewers should

be designed. Variations of it are commonly used in coeputing culvert
sizes and the size of Openings that must be provided in small bridges,

but it is essentially applicable to very small drainage areas and not

to the drainage areas involved in reservoirs constructed for flood control
or poser.

Any of the empirical eguations involving size alone or size

and other characteristics of the drainage area should be used only after

14
a considerable stuiy of the data on which it is besed to make certain
that it is applicable to the drainage basin in question, and then great
cure should be exercised in selecting the preper constants and coeffi-
cients.
legion: Flood of Record

, A knowledge of the largest flood that has occurred on the
stress being investigated is useful in many ester control studies and
should always be considered in attempting to fix the maximum flood that
is to be expected, even though there can be no fixed relation between
the size of these two floods. A record of stsges extending back from
50 to 75 years is sveilsble at one or more points on most of the large I
rivers in the eastern portion of the United States, while newspaper V
accounts and private diaries often.mention great floods which occurred
in the previous 25 or 50 years. If the greatest known flood occurred
prior to the period of continuous stage records, probably little is
known about it other than the year of its occurrence and its height above
low water or shove some other flood, and conflicting stories may make ~‘
it difficult to determine its height on present gages vith any degree
‘of accuracy. If the largest known flood hos occurred during the heriod
covered by gage records, however, its height, duration, and volume are
available, and such information will be exceedingly useful.

i In most cases, information will be looking as to the height
of the maxi-us flood of record at the particular point on the stream
where s voter control structure is preposed, but if the height is known
at points there the drainage area is not too different or at points

above and below the site, some estinste of the size of the flood at

the site may be made.

15

The liani Conservancy District used the maxinun flood of
record, that of 1913, as the basis for the design of their entire flood
control system. The discharge of the flood of 1915 was increased by
40 percent to obtsin the design flood used in planning and designing
flood control reservoirs and inproving the river channels at critical
points. This was done, however, only after a very careful and thorough
investigation of all stores which have occurred over eastern United
States in an attempt to determine the relation of the storm causing
the 1913 flood to the maximum storms that have occurred over.areas
subjected to similar meteorological conditions.

Similarly in other river basins, the greatest flood which has
occurred over a long period of record may be increased by 50 percent,
or perhaps more, to obtain some measure of the maximum flood to be
expected.

In addition to its use in determining crest discharge, the
hydrogreph of the maxinul flood of record is often useful along eith
hydrographs of other large floods of record in estinating the shape of
the maximum flood hydrograph. The greatest flood of record on a
drainage area was probably produced by the same type of storn that is
likely to produce the maximum flood to be expected on this drainage
area, and if this record’stornfiof the past were reasonably well centered
over the area, the shape of the resulting hydrogrsph should be quite
similar to_tbe shape of the hydrogreph of the maximum flood to be
expected. Thus, if the crest discharge of the maximum flood is fixed
by e series of studies, the daily discharges of the maximum flood of
record may be increased by the ratio of the two crest discharges to

obtain a reasonable hydrograph for the maximum flood.

16

In making use of the maximum flood of record, it should be
renembered that this flood is not the maximum that is reasonably to be
expected on the drainage area, nor does it bear any fixed relation to
the maximum flood, even though the flood records may cover e century or
more. Certain drainage basins have never experienced a flood comparable
to the great floods ehich have occurred on nearby basins having similar
flood-producing characteristics, probably due to nothing but the lack of
'rhyme and reason" thich characterizes stores and floods, and care cant
be exercised that this lack of great floods during the period of record
does not influence too greatly the final estimate of the maximum flood
reasonably to be expected on this drainage area.
Frequency ongast Floods at thei§ite

A nunber of methods have been developed for determining the
maximum flood to be expected in a given period of time from a mathema-
tical study of flood records at the site or on a nearby stream having
similar flood-producing characteristics. These methods are discussed
extensively in engineering literature and have been followed in the
design of nany mater control structures. The results obtained are to
be trusted no farther in the solution of the present problem than is
the maximum flood flow shown by the same records. Both are dependent
upon the length of record and upon the floods that happened to occur
during this period and are completely changed if different portions of
the record are used for the study.

The earliest of the frequency methods was that used by
I. I. Fuller in deveIOping bis equations mentioned previously. In
this eethod, a tabulation is made of the maximum flood occurring in
each year of the period of record, and the floods are listed in order

of eagnitude and numbered serially starting with the largest. Either

17
calendar or eater year may be used. The average annual flood is computed
and the size of each flood expressed in terms of this average. These
ratios are used in all following steps rather thin the flood flows
themselves, making the average annual flood the unit of flow. The
ratios are cumulated, starting with the highest, and each divided by
its serial number to obtain the average of all floods (eXpressed in terms
of the average annual) equal to or exceeding each serial number. The
total number of years represented by the record is divided by each
serial number to find the corresponding period of years. This procedure
gives the average of the floods (expressed in terms of the mean flood)
occurring in various periods of tine up to the length of the record.

When plotted on semi-logarithmic paper with ratio to the seen as ordinate
and tine in years as abscissa, the points fore a reasonably straight
line which can be extended to read the most probable maxi-us annunl flood
to be expected in any given period of time. The line must start with a
ratio to the averege of 1.0 for a time of one yesr since the average
annual flood must occur once every year on the average. The slepe of
the line is the coefficient of the log term in Inller's original equation
q = Q". (1 ./ 0.8 logm 1‘).

It is to be noted that the discharge obtained by this method
is the most probable noxious annual flood to be expected in any given
number of years and not the flood to be egualled or exceeded once in
this period of time. The latter conception is involved in most of the
other frequency methods.

Statistical methods have been applied to flood records in
many says to determine more accurately the relation between the size and
frequency of floods. A special coordinate paper h;s a vertical scale

which is either arithmetic or logarithmic for plotting ratio to the mean

g.’ " 18
annual flood, and a horizontal scale for plotting percent of time, the
horizontal graduations appearing close together near the center of the
sheet and increasingly farther apart near the edges in accordance with
the normal probability curve. If annual flood are studied as in the
Phaler method, the ratio of each flood to the average flood may'be
plotted against the corresponding percent of time during which this
flood was equaled or exceeded, the points felling along s reasonably
straight line which may be extended to read the size of the flood which
is likely to be equaled or exceeded a certain percent of the time or
once in any given number of years.

A method was developed by H. A. Foster for using the Pearsonian
frequency curves for mathematically extending such a frequency curve.
When the ratio of each flood to the average flood is tsken, the varia—
tion of each fro: the average was found and used to compute the
coefficient of variation and the coefficient of skew. Tables of factors
were prepared by Foster for various percentages of time and various
coefficients of skew, the factors being multiplied by the coefficient
of variation, added to unity, and plotted against the corresponding
percentages of time to obtain an extension to the frequency curve plotted
as described above. The extension should be a straight line or should
curve upvsrd or downward depending upon whether the original flood date
follows a normal frequency curve or is skewed.

Annusl floods are also used in a similar statistical method
presented by Allen Essen. His tables give rectors similar to those of
Pastor, each factor being lultiplied by the coefficient of variation and
added to unity before being plotted against the percent of time to which
each factor corresponds. Refinements and variations of these statistical

sethods have been proposed by later investigators, including Goodrich,

19
Blade, and Harris, while another variation was published by the
CalifOrnie Department of PUblic Works.

A method of sampling which is altogether different from the
'ennuel flood" system used in the methods just discussed may be called
the “basic stage“ system. Instead of considering only the largest flood
occurring in each calendar year and eliminating other floods which may
be only slightly smaller than the maximum of the year, this system
includes all floods above a certain besic stage, even though a large
nulber of floods may be included in certain years while other years
are not represented at all. The method does include, however, all
floods which are significant, since the besic stage may be chosen by
the investigator. The only difficulty is that a low basic stage
increases the length of the computations by increasing the number of
floods, and a too high basic stage does not give e fair sample.

In this method, all floods above the chosen basic stege are
tabulated in order of magnitude, and the size of each is plotted against
the frequency'with which each was equaled or exceeded. This frequency
is usually expressed in terms of the average number of occurrences per
one hundred years, or its reciprocal, the average interval between
occurrences. Plotting may be on any of the various forms of arithmetic,
logarithmic, or probability paper which will cause the points to fell
along e straight line or smooth curve which can be extended to obtain
the negnitude of less frequent floods. The curve should be extended by
eye, honever, end no attempt should be made to apply any of the proba-
bility methods, since the floods selected for the study occur at
irregular intervals of tine and do not represent the average sample

necessary for a true probability study.

20

As a variation of the basic stage method, all daily discharges
above the assumed basic stage might be used instead of only the crest
discharge of each flood. This is often referred to as the "complete
duration series” as contrasted to the "partial duration series" used
in the method described above. The occurrences are again listed in order
of Iagnitude and the percent of ties that each is equaled or exceeded is
calputed from the total number of days in the period of record. Since
Inny'lore items are involved in the complete duration series than in
the partial duration series, the procedure becomes laborious end the
items are usually grouped into classes and the midpoint of each Operated
upon as though it represented an item instead of a large number of
its-s. A smooth curve is drawn through the plotted points and extended
by eye. Bore again, theoretical probability methods cannot be used in
obtaining an.extension of the frequency curve since the original sample
does not represent a complete duration series but only the upper end of
such a series.

The relation between frequency end magnitude of past floods
as determined by the methods Just outlined does not entirely answer
the question of necessary spilleay capacity until a design frequency
is fixed. A structure vhose failure would involve only a minor property
loss eight be designed with e spillwey which would pass floods likely
to occur every twenty-five or fifty years, as it would be more economical
to rebuild than to support an expensive spillesy. then great damage
or loss of life would be involved in the failure of a water control
structure, it has been common practice to design the spillway for
floods of 500 to 10,000 year frecuency, frequency curves generally
been-ins so flat at the very high frequencies that there is little

increase in legnitude with frequency. The failure of a large storage

21

reservoir in a densely populated ere, hosever, would be so serious that
the spillwny should be ample to carry any flood which might occur
rather than s flood to which a definite freguency can be assigned. It
thus becoees important to know the limit of e freouency curve. This is
indeterminate in nest of the freguency methods, the flood magnitude
continuing to increase sonewhst sith frequency, thereby leaving in the
mind of the investigator considerable doubt as to the value of frequency
studies in the present problem. Further, when the amount that the
original flood data must be extrapolated (by eye or by one of the proba-
bility sethods) is compared with the length of the available record and
the effect of s slight change in the curvature of the extended line is
noted, the investigator becomes all the more dubious as to the value
of his results. Besides these uncertainties, it is to be remembered
that the available records on American streams are generally less than
100 years in length, and.nore often in the neighborhood of 25 years, a
length which is altogether too short to be of any value in predicting
the size of the SOC-year flood (s flood having one chance in 500 of being
equaled or exceeded in a certain year) as has been demonstrated so many
times by preparing frequency curves from different portions of the flood
record at a single gage and noting the variationlin size of the pre-
dicted BOO-year flood.
W

The 'unit graphI is o rather new tool of the hydraulician.
It node its first public appearance in 1952 in on article by'L. K. Sherman
end has since been investigated by many others with resulting improve-
rents and refine-ants but no basis changes. The principal assumption
involved is that s uniform rain of a unit duration over a drainage area

sill produce a hydrograph having a definite length, depending upzn the

22
run-off characteristics of the drainage basin, and having ordinates pro~
portional to the total volume of run-off included in the hydrogrsph. If
the ordinates of the hydrograph are modified by the ratio of the volume
corresponding to one inch of run-off to the volume under the hydrograph,
the resulting hydrograph vill be a "unit graph'-a hydrograph represent-
ing a unit volume of run-off from a uniform storm of unit duration, the
unit of storm duration generally being tsxen as one day to permit the
use of daily rainfall figures. Actually, it is impossible to find a
storm which is of uniform intensity over any sizeable drainage area
and which at the same time has a duration of exactly one day. Thus it
becomes necessary to compute unit graphs from a series of storms which
approach these reguirements, averaging the results to obtain a reasonably
accurate 'unit graph.”

The unit graph once having been determined for a given drain—
age basin, run-off coefficients may be computed from a series of storms
occurring during different seasons of the year and under various soil
conditions and applied to the unit graph to determine hydrographs of
run—off from any assumed rainfall. The daily amounts of rainfall are
multiplied by estimated run-off coefficients to obtain daily amounts of
runpoff. The ordinates of the unit graph are multiplied by each of
these amounts of daily run-off to obtain a series of overlapping hydro-
grephs of run-off. Ordinates for corresponding days on the different
hydrographs are added together to obtain total flov on each day, each
day's rainfall producing run-off over a number of days. In this manner
it is possible to compute the hydrograph produced by a single storm or
various combinations of storms.

In the actual use of the unit graph method, various refine~

seats are made in the method just outlined. Base flow is generally

23
deducted from total flow to arrive at flood run-off in computing the

unit graph, and similar estimates of base flow are added to the storm
run-off computed from the unit graph when determining stream flow from
rainfall. Elaborate studies may also be made to determine the runpoff
coefficients to be expected for reins of various intensities occurring
in different seasons of the year with various amounts of rain on pre-
ceding days and various ground-eater and soil conditions.

The unit graph provides a means of comquting run-off from
transposed and superimposed storms, and from a hypothetical storm
estimated to represent the maximum rainfall to be expected over the
drainage basin. It should be possible to predict the maximum rainfall
to be expected over a given area more accurately than it is possible to
predict the maximum flood to be expected from this same area. This is
partially due to the greater number of rainfall stations and to the
longer periods over which they have been Operating, but also to the fact
that great floods occur less freguentiy than great storms since each is
produced by a great storm (or a series of storms).having a center near
the center of the drainage area, a duration comparable to the period of
concentration of the area, and a distribution to produce coincidence
of floods from the tributary drainage areas, the soil and vegetation
conditions being Just right over the drainage area to produce a high
percentage of run—off from the rainfall. The maximum rainfall to be
expected may be estimated and the run-off cosputed by the unit graph
.eethod if the ground is assumed to be frozen or so saturated that the
run-off will be nearly 100 percent of the rainfall. In some areas, it
is necessary to consider a heavy covering of snow which will be melted
by the rains and added to the assumed maximum rainfall. Similarly,

large storms which have occurred over other areas having similar

24
netsrologicel characteristics may be trencpcsed to the area under study

and the runpoff hydrogreph computed for assumed runpoff coefficients, or
the eree night be subjected to various combinations of storms which may
be reasonably expected to follow each other.

The unit graph lethod of determining the mezinum flood to be
expected has several distinct advantages:

(1) The lexinun rainfall to be expected over the area can be
estileted eith some degree of accuracy, and e hydrogreph of run-off
can be computed for this rainfall and any desired assumption as to soil
saturation or smount of snow on the watershed.

(2) If the area is subject to seasonal stores, a maximum
store lay be determined for each season endused with the preper run-off
coefficients to determine the maximum flood to be expected in the
different seeeons when reservoirs are held at different levels.

(5) The unit graph is derived fro: run—off produced by storms
which have occurred over the area being investigated and so takes into
account the flood producing characteristics of the drainage area.

(4) The eethod giveefthe complete maxi-um hydrogreph, not
just the peak flow or the volume of run-off.-

There are also several disadvantages:

(1) It is difficult to determine e satisfactory unit graph,
especially when rainfall records are scanty and the drainage area is
so large that stores are neither of uniform intensity over the whole
area nor of e eellpdefined duration.

(2) The sethod seems to be limited to drainage areas less
than gum thousand square liles due to the difficulty of finding

uniform stores of short duration over the larger areas.

25

(5) The unit graph must be made from relatively small floods
during which the effect of channel storage is proportionally such loss
than would be expected during the maxinun flood.

(4) Application of the unit graph method requires that run—
off coefficients be estimated Just as in other methods.
Bali-us ecorded to of un—of over imilar Basins

The empirical equations discussed previously are based more upon
observed rates of run—off than upon theoretical considerations involving
rainfall, drainage area, and stream flow. Each equation represents the
conclusion reached by an investigator using the data available to his
ehich he considered applicable to a certain river basin or to a certain
section of the country. The observed maxi-um rates of run-off over
drainage areas of all sizes were probably plotted in terms of cubic feet
per second or cubic feet per second per square mile of contributing
drainage area against the drainage area in square miles. EnveIOping
curves could then be drawn to cover certain ranges in drainage area and
to include only certain types of drainage basins and basins subject to
certain types of stores. A factor of safety could be added if the
investigator thought it necessary, and the resulting enveloping curve
could be expressed as an equation involving drainage area andknaxinun
discharge, or length of the river, drainage area, and discharge if the
investigator cared to introduce another variable into his plotting.

It is often desirable to prepare similar diagrams to show the
esxilue observed rates of run—off in the drainage basin selected for a
ester control project and in adjacent basins having similar rainfall and
runpoff characteristics. Such a diagram presents to the eye a good
picture of the available information on flood flow, its variation with

sins of drainage area, and the agreement or lack of agreement between

26
flood floss determined from empirical countions, or other methods, and
the sari-us floss actually observed both in the area being studied and
in surrounding regions. The plotting may be either on plain coordinate
paper or on logarithmic paper. The later is preferable, since observed
sari-us flood flows generally fall so that a reasonably straight line
represents their upper limit on this paper, while many empirical flood
flow equations are of the exponential type and, therefore, represent
straight lines on logarithmic paper.

Considerable caution may be necessary in studying such a
diagram. If the available data on flood flees in scanty, the upper
linit indicated may be too low, and if the discharges assigned to past
floods were not carefully deternined, the entire diagram may be mis—
leading.

t o Run-off'flethod

The so-called “rational“ methods of computing flood flows
free rainfall are an adaptation and expansion of the method long used
in computing the flow for which city storm sewers should be designed.
The basic equation involved is that stated previously: Q = c i A, in
thick '1' is the contributing drainage area, 'i' is the intensity of
rainfall, 'C' is a coefficient eXpressing the ratio of runpoff to rain-
fall, and 'Q' is the resulting discharge. As originally used in sever
design studies and later adapted to slightly larger drainage areas, area
is expressed in acres, rainfall intensity in inches per hour, and dis-
charge in cubic feet per second.

The basic assumption involved in the use of this equation is
that maximum run-off will be obtained from a given drainage area when

it is subjected to a rainfall having a duration equal to the time of

27

concentration of the area, or the tine required for enter felling at
the outerloet lieite to reach the gaging station or dam cite at the lover
end of the area. This bringe up the corollary escunption that average
rainfall intensity o'er a given area is inversely prOportionnl to aura-x \\\2
tion eo that a higher intensity can be expected over a short interval E ‘N
of time than over a longer interval. Thus, in a drainage area of given
line and shape and having a given time of concentration, the maximum
flood rill generally occur when the entire area in contributing to the
flee at the lower end, and the runpoff from a storm ehoae duration is
leee thnn the tire of concentration eill be lees than the maximum
.'poeeihle runpoft since rain hee ceeeed to fell on the lower areas by the
tile rue—off fro- the farthest lieite of the drainage area reaches the
lower end. If the duration of the etore is greater than the time of
concentration, the entire drainage area may be contributing to the flee
at the lover end over a period of tine, but the intensity of the rainfall
and likeeiee the rate of run—off eould have been greater if the duration
eere somewhat lese.end.eore nearly equal to the time of concentration
of the drainage area.

It ehculd be noted, however, that theee atateeente ere tune
only in a general ray and that certain areas or.certein ctoree eay
iprcve to be exceptione. Droinnge areas ehich are long and narrow nay
hire Inch a long time of concentration that the intensity ie materially
redneed and greater flora are produced by a eore severe storm of shorter
duration covering only a portion of the drainage area. Large drainage
areas ney'have such a long tire of concentration that no single store
can be of sufficient duration, and the naxinue flood will be produced

by one etore over the headeatere and another store at a later tine over

the loner end of the basin. Direction of trowel of n etore may also

28
enter into the analysis of a large drainage area, a store which moves
dovnstreae so that it continually adds enter to the flood crest produc-
ing a greater flood than a more severe store which moves up the drainage
area so that rain falling in the headwaters tends to increase the dura-
tion of the flood crest without increasing its height.

The basic equation Q 3 C i A may be applied directly to snail
drainage areas, values of '0' and 'i' being obtained by estimate or
reference to a handbook. Such a procedure can hardly be called a
'Iethod' and properly belongs vith the empirical equations previously
given.

As an improvement over obtaining values of '0' and 'i' by
eetinate, various diagrams have been prepared to give these factors
for drainage areas of various sizes and in various locations through-
out the United States. Of the two factors, values of *C' can generally
be detersined vith less uncertainty. The winter or spring of the year
is the flood season in most sections of this country, and in consider—
in; an extreme flood, the ground say be eupected to be either frozen
or nearly saturated Iith eater at its beginning, thereby giving a run-
off coefficient of nearly 100 percent. In those sections where summer
stores such as West Indian hurricanes may produce the greatest rainfall,
the run-off coefficient is much more indeterminate. The sane is true of
very seall areas vhich nay be subjected to cloudbursts in any season
of the year. In his work on the rational run-off method, Herrill
Bernard has determined limiting values of the run-off coefficient through-
out the eastern portion of the United States, the values being a full
100 percent throughout the Ohio River Basin and decreasing towards the
southeast and northwest to a minimum of 60 percent in Florida and 50 per-

cent in the lestern plain states. These values are considered to

29

represent a frequency of 100 years and may be reduced for less frequent
stores.

The intensity of rainfall 'i‘ that must be applied to a
drainage area to produce the maximum flood is more difficult to deter-
sine since it involves duration, and this in turn involves the time of
concentration of the drainage area. Merrill Bernard has expressed the
intensity as 1 s 5131‘ in which '1" is the frequency and 't' is the time
of concentration. laps of the eastern portion of the United States
are prepared to show values of '1' for different localities, one chart
applying to duration periods from 5 to 60 ninutes and another for periods
from 60 to 1440 ninutes. Values of 'e' are shown in a similar fashion,
ehile values of '1' are shown as dependent upon location alone. The
time of concentration still reeains to be found, and Merrill Bernard
has prepared elaborate charts for estimating this from a knowledge of
the length and width of the drainage area, the length, slope, and size
of the principal channel, and a series of factors covering other charac-
teristics of the watershed.

This eethod of determining rainfall intensity and percent
run—off appears reasonable and should be superior to any estinete of
these coefficients, although values of these coefficients are dependent
upon records of somewhat limited length. The method, however, appears
too rigid to be practical in studying drainage areas of any size.

The most elaborate of the so-called “rational" methods of
coepnting flood flee divides the drainage area into a series of cones
so that rain falling any place within a sons will flow to the lower end
of that zone within a unit of time. The basic equation Q = c i A is

applied to each zone for each unit of time to obtain a hydrograph of

discharge free each zone which can be added to similar hydrographs from

50
the othor some, with prepor tho of travel allowanoo, to obtain tho
hwdrogroph ot any point on tho atroon. Tho oothod is oory laborious and
roquiroo oonoidorahlo dato on tho drainage area and its principal
Ito-om, but onoo tho booio intonation io authored, rainfall of any
ugnitndo and any duration on: Do applied to the drainage mo and tho
uniting hydropoph capntod without undue effort. Sono ndjnatnent
of tho hydrompha any to nooooury, hoomr, oinco the method does not
tuo into account the flattening offoct of channel storage, and thin
not to ootinatod or dotorainod by a prooooa of routing tho flood flooo
doon tho principal chomol.

tho first and most difficult my in the nothod 1: to divido
tho drninogo oroo into the noooaaary zones. Tho time required for water
to tron]. tron point to point on tho main channel and principal tributaries
io ounpntod tron o post flood ohooo protilo is known so that aux-ago
volooitioo can bo coopntod fro- knoon cross notional mad and did-
ohorgoo or in ootinatod by noon: of tho Manning oqnotion for Open ohannol
no- and tho “motion that tho hydraulic radial 1o oqunl to the flood
hoizht obovo loo tutor, or poaaibly throo quarters of this hoight, and
that tho olopo in tho mo as tho olopo of tho loo—outer profilo. (
Thou volocity and tin-of-trovol computations allow tho Iain rivor
and principal tributarioo to ho dividod into time zones, and it min-
nooooooryonlytooarrytnoao mootothoodgosofthobuin. Tho
oolo ootbodo oro appliod to all tho nincr tributarioo ohooo protiloo
oro aniloblo. volocitioo boing ootinatod in other omens thou pro-
moo aro not knoon.

Tho hit of tinoto bo and in laying out tho nonoo roqniroo
oaoo mlmuuon. loot rooordo of rainfall hovo been obtained from

51
gagee ehich were read daily, as automatic rainfall gages are too nee and
toe widely separated to furnish the necessary rainfall data on post
storne. lith rainfall records linitod to daily determinations, the sane
unit of time becomes convenient in studying run-off. unless tho drainage
area is sufficiently large, hoeever, to have a time of concentration of
eulothing like ten date, a eborter tine unit anet be need in order to
obtain enfficient points on the final discharge hydrogrnph, daily
rainfall figures being split up into the charter units of tine. A tine
unit of twelve hours, six houre, or even less any prove successful for
the elnller drainage areas, eith the obvious disadvantage that the
proooee bone-on eoro laborioue an the length of the tine unit is
doeroaood.

lith the tine sonee sketched on a nap of the drainage basin,
oechvnay’be planieotered to dotornino ita area. In the application of
a certain rainfall to the drainage area, runpoff coefficients nay be
chosen for each cone booed upon the season of the year, condition of the
ground, aaonnt of cover, etc. the investigator oonld choose coefficients
on the haeis of bio Jude-ant and knoeledgo of the run—off characteristics
of each.eoeo.

The investigator. may apply any rainfall Ihich he thinks
reasonable to the drainage basin, transpoeing past stores frca other
drainaao areas or preparing hypothetical rainfall distributions. Ho
any neon-o any run-off conditions which appear reasonable and may even
add in the effect of editing enoea in certain sense and frozen ground
in other tones, coopering the resulting hydrographs in tern: of croet
flov, total value, duration, and general shape. By neglecting the
tunnel! fro. eortein zones, he lay detornino the effect of a otornge
reservoir below those oonee or coepere the effecte of various poeeiblo

etorego reservoirs.

52

Other difficulties are generally encountered, hooever, ohich
prevent the probleo free being solved so siaply. Channel storage
usually operates to reduce and broaden flood create as the floods nove
downstreaa, a considerable volu-e of eater entering each length of river
before any increase in flee can take place out of the loner and of the
length. this values flows out of the length after the crest has
passed, thereby'incroaeing the duration and decreasing the height of
the flood. A siailar effect is produced by the difference in surface
slopes on the rising and falling sides of the flood save, the steeper
slope of the front of the save causing this enter to travel faster than
the enter at the crest, and the flatter slope at the rear of the Inve
causing that eater to travel closer than the ester at the crest.

in.olahorate routing scheme night he sot up to transfer the
ester along the rain channel fro- cone to none or free the south of one
tributary to the south of the next, or an estieate night he node as to
the probable crest reduction that is to be espected due to the above
factors and the reduction expressed as a percentage of the coaputod
Icreat discharge. A third possibility is to consider the rational
lethod as resulting in a hydrogrsph which is representative of the shape
and volnae of the final hydrograph but is indicative only of the crest
discharge} the crest discharge is estimated by some other nothod and the
cosputod hydrograph reduced to this crest discharge eithout change in
vein-e and without appreciable change in shape.

The rational run-off aothod appears to furnish a very flexible
and ieportant schene for estimating the maxinua.flood to be expected
free a drainage area shore rainfall is the principal factor in producing

floods. The nethod also provides the basic information necessary in

53
ccspoting flood hydrogreph which could result from any rainfall
occurring under any conditions of the drainage ores. A complete analysis
by this sethod, however, requires very complete data on the drainage
use and its principal stresss and very lengthy and laborious compute-
tions. Estimtes night be substituted for some of the cosputstions,
the rational method controlling the shape end values of the final
hydrogreph Ihile other methods ere used to fix the crest discharge.
The extensive study of the drsinegs ores required in the method has the
decided sdvsntege thst hasty conclusions cannot be reached end the
investigator must become thoroughly fanilisr vith all parts of the besin -
end its flood-producing charscteristics. The stores or hypotheticsl
reinfell distribution vhich are to be applied to the sres and the run-off
retes shich ere to he need require the careful judgment of one having
long experience in this type of sort and cannot be sttespted by an
sseteur. The rational. sethod is one of the honor ssthods of estimating
sexism flood hydrogrephs, and slang Iith the unit graph nethod, is
deserving of e considersbls amount of additional investigation.

W

W
Floods on the rennessee River at Chattanooga have been studied

by a great eany investigators. lriters of textbooks and articles on
eethods of estimating flood flow have often used the records of floods
at this point to illustrate their ideas, thile more recently the problem
has been studied in detail by the Corps of Engineers and the Tennessee
Valley’suthority in planning a systee of reservoirs for the Tennessee
River Valley and local flood protection works for the city of Chattanooga.

The city of Chattanooga is located on the Tennessee River in
the narrow portion of the Tennessee Valley Just above the gorge which
the river has cut through lhldren's Ridge and the connecting ridges
Ihioh divide the basin of the Tennessee River into two parts nearly
equal in area, sililar in shape, but altogether different in geology
and tepography. The conneroial and industrial portion of the city and
a sunber of the residential sections spread across the river's flood
plain, stile a substantial share of the residences are located on the
surrounding ridges far above the reach of any flood eaters. The city
has not suffered severely fron floods in recent years, but there is no
reason to suspect that future floods sill not equal or exceed in nag-
nitnde the large floods shich occurred 50 to 75 years ago. It has been
«tinted-t that a repetition at the present tine of the flood of 1867,
the sari-us of record, would result in a direct loss of $57,6000,000,
with an additional intangible loss shioh eight be even sore serious.
Thus, it is apparent that the city of Chattanooga needs flood protection-—
and a plan for its protection involves first a deter-ination of the
sexin- flood against which the city must be protected.

”louse Boon-ant lo. 91, 76 Congree, lat Session

55

A knotledge of the maxi-us flood to be expected one essential
for the design of Chickanauga Dan, nos nearly coupleted by the Temessee
Valley Authority across the Tennessee River a few niles above Chattanooga.
The deterlination of this flood at Chattanooga applies with slight sodi-
fioationito Chicks-auga Del.

- he record of floods at Chattanooga is one of the longest on
the Tennessee River so that flood estimates at other points along the
river nest be based to a certain extent upon this record. The location
of Chattanooga at the neural division between the upper and laser
basins of the Tennessee River allows the results of these flood studies
to be adjusted to apply for sons distance both above and bales the city;
Chattanooga thus acts as a good reference point to which floods on other
portions of the Tennessee River can.be related.

The problsa of deternining the maxi-us flood to be expected
on the Tennessee River at Chattanooga is therefore isportant in that it
affects the design of the flood protection sorts of the city, the
design of Chicks-Inga Del mediately above the city, and the deter-
nineties of the casinos flood to be expected sithin a considerable
distance along the Tennessee River.

BERKS! of the Bang

The drainage area of the Tennessee River at Chattanooga is
21,400 square ailes, or a little more than that of the Cuaberland River
at its south and a little less than the Susquehanna River at Harris-
burg. The basin is eoseshat elongated with a length of 260 miles, a
sexism width of 120 ailes, and an average sidth of about 80 miles. The
tepography is generally sountainous. lountain ranges of the Appalachian

systes rise to an average elevation of about 5000 feet on the south,
southeast, and north sides, shile the Gusherlsnd lounteins fora a 2000-

86
foot divide along the northeast side. Iithin the basin, the Great
Ssoky'lountains rise to elevations above 6000 feet in the southern
areas, and a series of generally parallel ridges extends over a large
portion of the resainder of the basin and reaches elevations between
1030 and 2000 feet. Exhibit 1 contains loco-toot contours shich shoe
the general tepographic features of the basin as sell as its shape and
drainage pattern.

The Tennessee River is forsed by the Junction of the French
Bread and Bolston Rivers a tee silos above Knoxville and about 190
river silos above Chattanooga to flee generally southwestsard parallel
to the Ohio and Cusherland Rivers. The French Broad River, vith a
generally'fannshaped basin, drains the eastern portion of the Great
Becky Iouatains. The Bolston River {love between straight, parallel
ridges throughout nest of its length, although it branches into the
lorth and South Forks to sake the upper portion of the basin fanpshaped.
The Little Tennessee and Hivessee Rivers enter the Tennessee River free
the south to drain east of the resaining portion of the Great hooky
lountains; the Clinch River parallels the Balaton River on the north to
enter the Tennessee River between the Little Tennessee and Hivassee
Rivers.

The surface soils of the Tennessee River basin above Chattanooga
are generally of a clayey nature, nearly ispervious to eater, shile the
rocks underlying the eastern portion of the basin are of the Pro—Cesbrian
era and therefore ispervious and lacking in the sinks, caves, and
solution Channels shich are prevalent in the shale and lisestone rocks
of the lississippian era vhich cover the resainder of this basin. a
shell ascent of surface storage is provided by the linestone sinks, but
there are no lakes and ssenps to furnish surface-storage comparable to
that found in seat river basins.

‘ 57

The average annual rainfall averages about 52 inches over the
basin, fro- a sininul of around 40 inches at stations in the sheltered
portion of the French Broad Valley to a naxisun of around 80 inches on
the exposed peaks of the higher nountein ridges foreing the southeastern
boundary of the French Broad area. Exhibit 2 shove the isohyetals.
depicting variations in the 50—year mean annual rainfall over the basin.

The area is subject to cyclonic stores of the type that cross
the eastern portion of the United States from the scuthsest to the
northeast, and sons portions of the basin say be subject to intese
stores caning free the Gulf of lexico. The subject of stores and store
rainfall sill be discussed in sore detail in connection with the rational
sethod of estimating flood runoff.
Available gydrolggic pets

The record of stages of the Tennessee River at Chattanooga is
prwtlcally couplets from April 1, 1874, then the first staff gage ens
constructed by the U. 3. Signal Service, the predecessor of the present
8. 8. leather Bureau. Gage records are not available for the flood of
1867, the greatest of record, but the height of this flood has been
deter-ined quite accurately fro- flood lurks, and the general shape of the
hydrogranh has been pieced together free nevepeper descriotione of the
flood. ureat floods ere knoen to have occurred in 1826 and 1847, but
neither of these rivaled that of 1867, and it is doubtful if any other
' great floods occurred in the period of legendary records, 1826 to 1867.
Exhibit 3 gives a sulnary of important data relative to the vsrious
river gages Operated at Chattanooga; Exhibit 4 shove the principal
floods of record.

Discharges of the Tennessee River at Chattanooga have been

lessered by the Corps of Engineers and by the B. 8. Geological Survey,

58

but the stage-discharge relation has been complicated by backwater from
holes Bar Dan constructed by private interests in 1915, at a point
about 85 silos holes the city. Various stage-discharge curves have
been drain and extended to indicate the discharges reached by high
floods of the past and the stages that night he reached by future floods.
These rating curves have been revieeed in the Flood Control Section of
the Tennessee Valley Authoritqiand a slightly different curve has been
prepared ehich agrees reasonably sell eith all discharge neasureeents,
eith the discharge indicated for great floods free a study of these
floods at other points in the Tennessee River basin, and eith backwater
curves computed throughout the length of the Tennessee River. The dis-
charges indicated by our rating curve are felt to be reasonably accurate
over the cenplete range of flee to be expected, one sectiOn of the
curve applying to conditions prior to the construction of Hales Bar Dee
and another section to conditions after the construction of this den.

stage records have been kept at a number (f other points on the
Tennessee River besides Chattanooga. The following tabulation shove

the dots of establish-out of these early gages, the river mile, and the

drainage area.

m ve e Mimge grea Established
:0mm. 98.‘ 58,500 “to“? 1. 1875
Rivertcn . 226.5 31,560 lay 18, 1891
Florence 256.6 50,810 levenber 7, 1871
Decatur 504.8 28,900 October 1, 1875
Bridgeport d14.d 22,600 larch 25, 1892
Chattanooga - 464.2 21,400 April 1, 1874
Kingston 568.2 12,500 October 1. 1874
Knoxville 647.2 8,900 January 1, 1875

Rain gages have been Operated at many points in the Tennessee
[hiver'basin by the U. 8. leather Bureau, sons of these records extending
farther back than those of the river gages. 'loet of the rain gages,

59
however, have been located in the larger towns along the rivers and
very for records are available free the higher and sore inaccessible
areas. This lakes it difficult to compute average rainfall over the
basin for any store or any period of tine. [any additional gages of
the automatic type have been installed by the Tennessee Valley Authority
in the regions having a deficiency of gages; Exhibit 5 shoes the location
of both rain gages and stress gagee in the basin above Chattanooga.
1 tion cal us

lbpirical flood flow equations are of little assistance in
estisating the saxieue flood to be expected at Chattanooga. while the
elieatological and flood-producing characteristics of the drainage area
appear quite sinilar to those of other areas in eastern.United States,
the drainage area of 21,400 square silos is considerably larger than the
areas free shich flood flow equations have been developed. Consequently,
lost of these equations lust be extrapolated beyond the limits of the
data free thich they ears derived-a process ehich is very questionablev
sith any data and very dangerous in this case. 1

may s equations sere derived, fros flood no. data for
drainage areas less than 5000 square silos on the flohaek River, and an
-attespt to extrapolate these equations to the drainage area of the
Tennessee River at Chattanooga gives the follceing absurd results:

Rare floods, o s 127 l 1.4 a 284,000 cfs
I 570
Occasional floods, Q s % I 20 a 470,000 cfs

The lurphy equation developed for areas up to 10,000 square

silos in northeastern United States gives

9:22.129 {153570,000cfs
s/szo

40

The Pinningequstion for use England streams gives
Q I: 800 .5/6 I! 812,000 efs, while the very eisilar Dickens equstion
for the Central Provinces of India gives 0 .. 500 a} = 885,000 cfs.

the Ieyere equation for extra-e floods as modified by Jervis
gives 0 a 10,000 xi = 1,460,000 cfs.

the Fuller equation with '0' estisated at 70 and '7' taken
as 1000 years gives 0 : 0 10-3 (1 l 0.0 iogm r) : 590,000 cfs.

The general width foreula of C. R. Pettis with a 1-day rain-
fall of 7 inches and an average basin width of 47 miles gives s probable
100-1»:- flood of 0 : 480 I I1-25 = 415,000 01‘s.

h rv te of unoff

The record of stages of the Tennessee River at Chattanooga
dates back less than 70 years, while the entire period over vhich there
are actual or legendary records of floods is less than tvice this
length. The most reliable nethod of supplementing this record in
sttespting to fix the maxi-us expected flood is through the collection
and comparison of the rates of runoff experienced during major floods
which have occurred in drainage areas subjected to similar stores and
having einilsr runoff characteristics.

A etudy'of the stores which have produced the major floods of
reeerd in the eastern portion of the United States indicates that these
eey be divided into two general classes: those of the cyclonic type
which love across the United States fro. vest to east or free the
southwest to the northeast, and the‘weat Indian hurricanes which move up
late the United States free the south or southeast. The sountain ridges
thich surround the basin of the Tennessee River above Chattanooga are
high enough to force any lest Indian hurricanes which may travel this far

inland to drop nest of their soisture along the edge of the basin rather

41
than over the basin itself in a general store of the duration required
to produce the anxious flood at Chattanooga. This leaves only stores
of the cyclonic type to be considered as being capable of entering this
basin and producing the naxisun flood and eliminates from our considera-
tion the floods produced along the Gulf and Atlantic Coasts by lest
Indian hurricanes. The remaining portion of eastern United States is
subject to cyclonic stores and is therefore comparable to the Tennessee
Valley in this respect.

The basin of the Tennessee River above Chattanooga is generally
sountainous vith steep slopes, ispervious soils and rocks, and lack of
surface storage. ‘l0et of the rivers draining the Appalachian flountains
are siailar in this respect. Thus, the flood-producing characteristics
of the Connecticut, Susquehanna, Delaware, Potonac, Hudson, Cumberland,
and upper Ohio Rivers are generally similar to those of the Tennessee
River above Chattanooga, and the casinos rates of runoff from these
river basins nay sell be studied in attempting to fix the maxi-us flood
to be expected at Chattanooga.

in examination of the saxinun recorded rates of runoff from
these rivers draining the Appalachian ayates discloses that practically
all have been produced by one of the following seven great stores:

lay aleune l, 1889-Susquehanna Basin '

larch 25—27, lOlB—-Ohio River Basin

lovesber 2-5, 192?-Canneoticut River Basin

larch 21-23, 1929-—Tenneesee and Culberland River Basins

July 7—8, 1955-Upper Susquehanna River Basin

larch 1e—22, lDSfio-Upper Ohio, Susquehanna and let England

Jslmary 6-25, 1957-Lover Ohio River Basins

The saxisus rates of runoff produced by these stores in the
various basins sere carefully tabulated from the records of the

U. 8. Geological Survey and other reliable sources. The entire tabulation

42
fills 62 sheets and is too voluninous to include here, but a sanple sheet
is reproduced in Enhibit 5.

To provide a visual comparison of these various rates of
runoff, runoff in cubic feet per second per square mile of drainage
area sas plotted against drainage area on log-log paper, different
eyebcls being used for the different floods. The resulting diagram is
reproduced in Exhibit 7. It is noted that the highest points on the
diagran define rather precisely a line passing through a runoff rate
of 50 cfs per square sile for a drainage area of 10,000 square silos
and 500 cfs per square sile for a drainage area of 100 square miles,
this line having the equation o . ffi in which '0- is the runoff
rate in ofs per square sile,and '1' is the drainage area in square
siles.

A sisilar diagras shoeing observed sexies- rates of runoff
eas prepared free preliminary studies made in 1954.55 in connection
lith the detersination of the naxinul floods to be expected at various
points in the Tennessee River basin. These studies sore reviesed by a
special board of consultants conposed of Harrison P. Eddy, Ivan E. Bonk,
Gerard R. letthes, and Daniel I. lead, and a report was submitted by'
this Board under date of lay 29, 1958. In this report the Board
recon-ended the above relation between sari-us flood rates and drainage
area for general application to drainage areas in the Tennessee River
basin of sore than 500 square niles. In making this recommendation,
hosever, the Board of Consultants recognised that in dealing sith
particular drainage areas “each problen should be studied in detail
eith the exercise of Judgnent in the light of the local conditions.”

For the Tennessee River at Chattanooga sith a drainage area
of 21,400 square miles, this represents a runoff rate of 54 cfs per
square nile, or a total discharge of 750,000 cfs.

as

It is to be particularly noted that the enveloping curve or
lisiting line shown on Exhibit 7 is supported in the vicinity of the
drainage area at Chattanooga by a point representing a discharge rate
of 50.7 cfs per square lilo for a drainage area of 24,100 square silos,
the eylbol indicating that this point belongs to the store of harsh 14—
22, 1956, covering the northeastern part of the United states. actually,
this point refers to the Susquehanna River at Harrisburg. The flood-
produoing characteristics of this basin are similar to those of the
Tennessee River above Chattanooga, as the watershed of each is principally
sithin.the Appalachian.lonntaine with steep slopes, loe infiltration
loss, and.negligible surface storage. A second point is also noted at
the sale drainage area but with a runoff rate of 29.0 cfs per square
eile. This point represents the 1869 flood on the Susquehanna River
at.flerrisburg. A study of this flood and the store shich caused it
indicates that the greater portion of the rain fell in slightly over
24 hours-too short a period fer the entire drainage area to be contri-
buting to the flood crest at aarrisbnrg. This ieeediately suggests that
the 1889 flood is not the sari-us to be expected, a conclusion which
has been.verified by the 1956 flood.
new

A long record of past occurrences is often the best basis
for predicting future events, but care oust be exercised to make sure
that.the available seeple is a fair-sample and that the data which it
presents is reasonably accurate. A record of floods at a single gaging
station seldom, if ever, nests both of these requirements, and the record
of the Tennessee River at Chattanooga is no exception. This is not
seriously objectionable if past floods are to be used in formulating

44
a reservoir operation plan or in sstisating how frequently a highssy
say be flooded, but it becoees very serious if a record having a length
of less than a hundred years is to be the basis for fixing the size of
the sari-us flood for ehich a water control structure sust be designed-—
a flood which night occur once in 500, 1000, or 10,000 years. Accordingly,
the record of past floods at Chattanooga was not used as the basis of a
statistical deternination of the nuisus flood to be expected, but
studies were node to shoe the great inadequacy of this record for any
such purpose.

A gage has been Operated on the Tennessee River at Chattanooga
since 1874, and daily readings have been published by the 0. 6. Feather
Bureau over the entire period. Discharge seesuresents were first node
by the Corps of Engineers and later by the U. 8. Geological Survey,
the latter organisation operating a recording gage for sany years and
publishing daily discharges over the greater portion of the entire
period of record. Although a nnsber of gages have been operated at
and near Chattanooga by both the leather Bureau and the Geological

Survey, the Weather Bureau gages have been located relatively close
together and have been set to practically the ease datu- sc that gage
heights are generally coepsrsble over the‘period.of record. Binilarly,
although the discharges published by the.Geologica1 Survey have been
deter-iced free gages located at Hales Bar and Bridgeport as sell as
at Chattanooga, the drainage area at the various gages is not materially
different and the discharge records are generally conparable. The
discharges of the great floods which have occurred at Chattanooga, how»
ever, are quite uncertain. lost of the Chattanooga discharge seasure-
sents were node dhring recent years, and the highest recent flood is

that of 1917 with a discharge of about 541,000 cfs as coepared eith an

45
estineted discharge of 459,000 cfs for the naxinun flood of record,
that of 1867. Further, the anxious measured discharge is only 276,000
cfs, and only 7 discharge measurements have been ends for flows exceed-
lug 200,000 etc. The discharge neasurenenta themselves may be in error
by appreciable anounts, especially those made during great floods when
the river velocities are high, the weather is cold and rainy, and the
enter is filled with floating and subeorged debris of all sorts which
is apt to dosage or carry away the measuring equipment. The stage-
discharge relation probably has changed an appreciable amount between
the early floods of record and the time of the discharge measurements
upon which flood discharges must be based, stages at Chattanooga being
increased somewhat due to the construction of Hales Bar Dan 55 silos
below the city and to the encroachment of the city onto the flood plain
forecrly occupied by the river. Stages say have been reduced somewhat
due to leproved align-ant through the city and clearing of the wooded
portions of the overflow area for none distance above and below the city.
The discharges published by the U. 8. Geological Survey are taken from
stage-discharge relations prepared from ties to tine froe the available
discharge eeasurenents and are not all taken from the sons rating curve.
This procedure may take into account some of the major changes in the
stage-discharge relation, but the resulting discharges are apt to be
'soueehat unreliable at high stages because the rating curves are drawn
fro- insufficient discharge measurements. Consequently, for the studies
ends by the Tennessee Valley Authority, a particular rating curve was
adapted for conditions prior to the construction of Hales Bar One and
another curve for conditions after the construction of the don, the

tee curves uniting at a discharge of 540,000 cfs and renaining coinci-

dent at all higher flees. These curves are based upon all available

46
discharge eeasurelents, upon a consideration of the crest discharge of
pest floods ss indicated by s study of these floods st other points in
the Tennessee liver besin, end upon backwater curves coeputed throughout
the length of the Tennessee River for both natural and reservoir condi—
ticns.

Although gages have been operated at and near Chattanooga by
both the U. 3. leather Buearu and the U. 3. Geological Survey and studied
by both this agency and the renneesee Volley Authority, it eust be
recognised that the estiseted discharges of the higher floods ssy'be in
error'by as such as 10 percent, and this possible error sust be kept in
sind when u» dats 1- used in studying pest floods and in predicting
possible future floods.

the length of record at Chattanooga which is available for
eating couplets frequency studies is less than 70 years, although con-
sidersble-inforestion.has been gathered pertaining to the flood of
1.67, and high floods are known to have occurred in 1826 and 1847. This
is hardly sufficient inforeetion to cake an estisate of the 10,000-yesr
flood or even the 600~yesr flood, especially shen a study of the data
shows that it is not a 'fair couple.“

to shes the fallacy'of atteepting to extrapolate flood dis-
charges fro-.a frequency curve based on the record of past floods at
Chattanooga, a series of frequency'curres were prepared using different
portions of the flood record and assuming that these frequency studies
night have been ssde st various times using the records svsilable at
those tiles. Inhibit 4 shows the magnitude of the past floods st

“cogs; the top dingran gives a picture of the distribution by
3 loser portion shows the distribution by seasons. It is

47
noted that the three greatest floods of record, those of 1867, 1875,
and 1886, all occurred in a period of 20 years, while the 20-year period
fhllowing the flood of 1917 does not contain a single flood comparable
to any of these and only a few floods amounting to more than 50 percent
of the highest. Using the basic stage method and considering all winter
floods having a discharge of 100,000 cfs or greater, the resulting 500-

and lOOO-year floods are as follows:
zerigg gongidereg Eggber of Iparg 500-135; Flood ‘lgggzzgg;;[;ggg

1867 to 1886 20 750,000 cfs 840,000 cfs
1887 to 1955 49 421,000 450,000
1867 to 1955 69 545,000 596,000

This is a variation of 78 percent for the SOD-year flood and 87 percent
for the loco-year flood. The discrepancy doubtless would increase to
even greater figures if an attelpt were node to determine the naxisun
flood to be expected. The various frequency curves and their extensions
.r. shown on Exhibit 8.

The record of floods on the Tennessee River at Chattanooga
has been used by several writers to illustrate the methods they have
proposed for estinating the nagnitude of the maxi-us flood to be expected
or the sagnitude of the fhod corresponding to any frequency. A cospari-
son of these figures is interesting in that it shows the great variation
in the results obtained by the different investigators. It is to be
noted, however, that all of this variation is not due to the method used
in handling the data, but to the data that was used-the period of record
selected and the discharges adopted for the various floods. The various
frequency curves are reproduced on Exhibits 9 to 15, and the resulting
floods of 500,1000, and 10,000byear frequencies are shown together with

the reference, sethod, and period of record used. All the frequency

48
curves takenfroeieter supplyraperWlu-ebssedcnthe saseperiod
of record and the ease discharge data so that variations in the results
are wholly'due to the eethod used. In the case of the 500+year flood,
the variation is free 56l,000 to 405,000 cfs, or about 12 percent,
while in the case of the 10,000-year flood, the variation is free
580,000 to 400,000 cfs, or about 25 percent. The 25 percent variation
is reserhably seall considering that a record of less than 100 years
has been extrapolated to 10,000 years, but any sense of accuracy of the
results is short-lived when the rescinded- of the table is exaeined. As
previously explained, the frequency curves shown on Inhibit 8 are based
on different periods in the flood record at Chattanooga and show a
variation in the-selves of nearly 90 percent for the lOOO-yeer flood.
The eaxilue variation considering different periods of record and
different detereinations of flood floss is free the 840,000 cfs shown
on Inhibit 8 for the lOOOuyear flood using the period 1867 to 1886, to
the 375,000 cfs shown on Ixhibit n m the lOOO-year flood es deter-ined
by the Goodrich.Type V eethcd using the period 1875 to 1881, and dis-
charges as published by the U. 8. Geological Burvey. This variation
aaouats to about 150 percent.

further evidence of the erratic results likely to be obtained
by using frequency eethods based on relatively short records and dis-
charges ehioh are of questionable accuracy is sheen on Exhibit 14.
Here the sagnitude of the 500-year flood expressed in cfs per square
ails of drainage area is plotted against drainage area, the frequencies
having been coeputed by the Corps of lhgineers using the Goodrich.eethod.

The points scatter rather widely about a eean line having a slope which

~ * e uare root of
indicates that the discharge varies in prepoftion to th. 'q

49
thedreinegeerea. Thereshouldbesosescatteringofthepointsdue
tethedifferenceintheflcodprodecingeherecteristiceofthedifferent
draieagebseies, betthevarietion of several hundred percent see-etc
indicate thetthsfrequneynetbod oennotbeappliedtorecordsesshort
elddisnhergeseslaoerteieasthoseinthe'fenneseeetivsrbasin.

nigchaggc in 19% fig 1_
5 1000 000

m1 him new range 10.8.1: nae nit
. 9 Hydro-Electric Handbook Basic Stags 445 470 570
Greager a Justin, 1927 Iearly flood 485 480 525
10 'eter supply r eper 771 Foster type I 1875—1951 571 878 sec
111 ' 405 421 476 -
Hem ' 405 424 480
ll ' Goodrich type '1 ' 871 876 400
II ' 680 598 428
V ' 361 578 594
. ll ' Slade " 416 428 474
16 an. 328 Goodrich 1876-1827 410
Corps of Engineers
8 as leeerds ‘ roster 1867-1866 750 850
1887-1955 405 450
1867-1855 511 586
W

These-celled'rstional'ssthodofsstieetingrunofffr-rain-
fe‘llwasdessribedefisceeofitsverietionswerediscusseduder
'Cc-cnlethods.' Thisssthodhasbeennsedtoaccnsiderableutentby
theMeIseeVelleysnthcritwinestieetingthe-xi-Iefloodthetsight
mathemhmetmetme. ”previouslyuplaised,
thedreinegeercsisdividsdintoeseriesofsenesscthatrainfelling
esypleeewithiesscnewillfleetotheloesreedcfthatsonesithine
nitoftiee. Thebesioequaticnqzcisissppliedtoeachsenefor
subtitoftiaetoobteinehydrographofdischargefroleechsueahich
celbeeddedtesililerhadregrephsfroeetherscneswithprcpertiseof
travelelleeeneetoobteisthehydregrsphetenypointcnthestreee.
Boeeedjustesntlustbesadeintheresaltinghydrogrephsincsthessthod
deesaeteusidsrtheflattuingeffectcfchaenelsterageuponflced

50
waves,orelsesneleberateroutingpsoeessenetbearrsngedtctrensfer
theflewdontheprincipalstreaefrcescnetosenecsatleesthetwen
principaltribttsries. AtteeptingtorcutetheflmdmtheTennessee
liver and its principal tributaries is both laborious and of meertain
eesussyetthepresentstagecfdevalopentelcngthisline. Therefore,
erestdisshergesdetsseinedbythereticaelsethodwereredeeedfcrthe
effectef channel storage. The rational eethod furnishes thevolueef
r-effsndtheshapecfiqdrcgraph,whilethepreviouelydiscesseddieun
stead—observesrneffratesfarnishestheeexieusepectedflow.

’Theweiteftisetobeusedindividingthebesinintosones
wesfirsttahuasuhonrs,eosteftheavailablerainfallfiguresbeing
for24-hourperieds. hates-studies, hosever, indicated thattheresulting
Wereutulleeoughdefieed,destcaninseffioientsnebsrcf
points, ands-howe- senesvere substituted. The floodofneeesber, 1882,
furnishedthebasiedatefcreeepstingtiseoftravel,asthisflocdwes
efnffisisatsagnitesstogivereeseaablyhighvelositiesendeemed
recentlyeneeghtoheveitsprcfileenddisehargesrethsrwelldefieed.
ummmmmmmfiorm,mmmuum
o-putedbyesenscftheleuingeqnetionfercpenehennelflow. The
river syst. was divided into reaches sufficiutly short so that points
achtmalsoftravelsloegtheriver systuabcvemtteneega
couldbeinterpolated. Theses-hmtraveldisteneeswereoontinnedepthe
Tonesseeliver, itsprineipeltributsries,andalltheainortribetsries
enshichtheeeeesssrydatewesavailable. sis-hourtreveldistaneesvere
estieetedoverther-einingdistaeeetcthebesinboundary. fiepoints
detaliudinthismdefineliaessieilarteocntcursmsptthatthey
rmesntuseemehettanoogeretherthenfeetsbovesdettnplene.
MibitlbshovsthsresultingO—hourtisecftrevelsonesabonmttenooge.

51

Theseus-floodtobeespestedatChattanoegawillbeps-odnoed
hyastcsehavingednretionabcetequeltetheooneentreticnpericdcf
thedrainageereyorahutfivedeys. Alongerstorewouldhaveescse-
whatlessintensity,whileeshorterstcrewouldnctceuseallthe
dreieegeereetoeutribntetothsfloodsrest. loettenptwessadeto
mscnttheesxisI-possihlereinfalltebeexpectedovsrthedraisage
eree,bntgreetstcrescfthepsstweretransposedoverthebesinendthe
resulting ths esaputed. This required a detailed study of pest
stesuevertheT-nesseeliverbesinendadjeoentareastodetereineehat
typesefstoreseightbeexpsstedtcoecurinthishasinendtheesgnitude
eadprebeblesessencfeceurrenoecfthesestorse.

Theseeeoeeldistributioeoffleodsatdhsttanocgasspioturedon
Inhibit4indicetesthstthe0hettsneogefloodseeeonesteadsfrceabout
thesiddleefneeewhertethesiddleofspril,thegreetestfloods
occurringinlarehendthefirstweehcfspril. Thisseessnaldistributicn
isupleieedtoeoertainexteeteedsnppcrtedbythebesicsetecrologissl
deteshcweonllhibitlc. Itisoheervedthsttheeveregeecnthlyreinfell
udetsrsinedfreeelergennhsrofreprseeatativestetionsinthedrsinsge
areaaboveChettenoogahevingWD-riodscfreccrdisgreeteetinlereh
bateeerlyssgrestinnec-ber,l\ny,endsugnst,therebeingnogreetor
wellndefinsdseesoualvariationinrsiafall. Thesisilareurveshowing
everegeruoffatmattseoogaishighinlsnusryandrebruerybutresohesa
definitsnxinielereh,dremingtolessthensthirdcfthisaecnt
duriegthes-ersonths. Themvecfaverageperoentmoffisstill
serestrihiegwitheesxis-efeboutmperoutduriuthewinterscnths
chsnuery,lebruary,hrch,endspril,sndseinieueofslightlyover

”percentdurisgthes-ernonthsofl‘aly, August, 8epteeber,and0ctcber.

It
Thislewpsreentrenoffduingthesu-erecnthsasecepsredtothewinter
soathsappeerelegicalwhentheothercervescnthischsrtareeonsidered.
Thedayisseentebeoverwperesntlsnger(sunrisetemset)ieJ-e
thenatthefirstcftheyeer,theesrveofaveragetenpereturebeingvery
sieuarwithseexieuinluly. Lengdeysandhight—pereturesprceote
uyiuefthescilwhichthenrequireep-eeterreiefalltoprodneeset‘m-
atieeandhighrstescfrueeff.

hr the present purpose, stores who grouped into three classes:
winterstoraseftheeyeloeiotype, Ieetlndianhurricanes,andsoeser
mm. or eleudbursts. The last named stores are umpanied by
veryinteeeepreeipitetionbntthedurationisshorteedthesreacevered
sesllsethetthistypecfstonceuldnctpossiblyprodncetheaexin-
floodtoheexpeetedfreeslargedrainagearea. lsehyetaleapsoftgpicel
stmscfthistypewhichhsveoccurredinthei'ennessee'alleyregice
areshesncnlmbitl‘l. Itisnotedthatthemyintenserainfall
ereascevercnlyafewsqaeresileswhileesshstorseeversatoteleres
lesstheeahnadredsquareeiles.

lestlndienhsrrieeaeseretropicslstereswhichoriginatenesr
theeqnatoriethevisinityoftheleetlndiestotravelscrthwsrdthroegh
the0elfoflexicesndaerossthescuthsrestates,generellycurving
towerdstheitlantisdoeensndseldoeoseingveryferinland. They
Winteeupruipitationcvershcrtperiodsoftisesedoverrelatively
lergesreas,tatgenerallyoecurinthelatesueeersonthsshenrunoffis
.a-allpescntageofrainfell. Matthistypesrelihelyto
producenxisuflocdsintheereessubjesttosuehstorss. Thereis
easiderabledoubt,however, thatsuchastorlsculdtrevelthisfar
inlendanderossthsecantainseurroundingtherennesseeliverbseinand

stillretainsnffisienteoistnretoprcdnceeesxil-fleodonelesge

‘8
drainage a... nu stcse of July‘l4-18, 191s, which We» greatest
flsodcfreeordontheupperportiasoftherrenchlreadliverwascf
thistype. hemmannsreoordedsttwepointsiniorthceroline
entheeesternbeundaryoftheTennesseeniverbesin,thesepointebeing
mappedtesidescfaeeuetaingepabeuteidwaybetweenmndfather
Imteinwithanelevetionoffl‘dendleontlitchellwithenelmtion
arm. Altepessonthseastsleperecordedareinfallofud71nohes
ineperiodalittleeverflhcurs,whilesltepasslnnonthewestslcpe
had22.flinehes. lainfallstatiensesretoofevinthisareetogivea
scepletepictursoftheietensityendduretionofthisstore,butae
isohyetaleapwssprepsredbytheliuidcnservaneyMstrietisecsnectien
withtheirsterestudiesandisrepscdeeedelengwithotherstcreson
mm as. The chain of the Tennessee Valley Authoriw indicate that
aseseshetdifferentrainfallpettsreoanbededessdfrcsthesseerain-
falldata;thereslltsereshcncnlxhibit18. oonsidsrstioe ishere
given to the effect ef topography on, rainfall, resulting in a greatly
rednesdvolueefreiefellwithintheTennesseehiverbasin. Gaging
statieasieepereticeat thistieewerefewenddischarge estieatesncne
tocucueu,bntruoffstudiescnthelrushlrosdliveriadiutethat
lestofflsereinnsthsvefellnonthssoutheastsideof‘thescunteins
ratherthaninthereenesseeniverbesieonthenorthwestsideefthe
mum. llhibitlfishoesthepethscflcwwessureereeseeeowpewing
great hurl-issues which have approached 'cr crossed the Tennessee liver
bssinwhiledaadedstripsindicatethesreesofgreetestrsinfall.

' The lost ilportent stores as far as Chattanooga is concerned
aretheseefflzecyolouietypewhiohtrevelscrossthiscmtryduring
thewintersesscninagenerellywesttoeestersoutheesttoncrtheest
directia. StonscfthistypeheveprcMedsestofthegrestfloods

ormummmmmumwmwm
umamrflmmmummnmnmu»
MQWT-Iuloollm. muttoshoosthepaths Iona-db:
smdmusmsmMmrmrmsmuvu-basmmu
whitaumlupesmeamuammspomurwmanas
«amummmmwmummm
fennesseeliverhash. west-u, ”.mushoeisehfitalupsot
themtetensofmmmudltetes,1nolmn‘theston “MM-16,
amass-uumsmuorwqumcwp. Inordoroegetahettu
nature or the Wave intensity or the various pat. stern om
ever Mono of eastern out-s Itatu nbjoot to meteorological audition.
stlmroethuettheremessee'mey,aser1esofuse-area-depth
mmmtmamutmmmnsmmmuumum
shrafiuottws,ot8au4¢ays,andotidsysormr.

mumamsuaotaboutbmsmotpeatest
mummtmmmmmmmstm
huseonenfiratlenuneoraboutthism,butn1slenum
“Motsughtlypeatuowleumuon. hummus.-
srea-depihmsferosnda-dsypuiods,1t1saetedthsttorthe
mummmp(u,soosqmamu),wa—aqm
«Immunflufisnmramtanorsboutmiuheshm
lostsevere;theHsystonetOetohez-H,19101s1ess1nteosebyon1y
halt-aim. mnflstonmmdapomonoftheMesseeum
mummhnduoloutothobsunthattmuutuoquem
mustanpeuuonorMsstorn-ightreasonauybeupeeteduthis
ares. W,th1sstmooourredmmtheearlypanotthenood
season at chattsnooga, shilethelOlOstor-oooorred ontsideths

Chattanoonneodseasoa,,mtheu¢h1talsooomrede1m:areamble
mumrenesseeumhasin.‘

a I!
WWMWWI-depthmhrms
«Ida’s-aim.tetsseenthsttoradrshsgemathesiuo!
memmemcneheI—dsystmotflly 0-10, 1916. eithsn
mahfauofsbootlbtnohee,1staraboveauothere. hub“ 10,
m.ehoueeh1sembbeeieetm1anhmimeehiothuflh
mmmmumumpmnoruuoummmm
“tether-onus” lens basin. mmsmum11.u.1aoe.
udthi—hyetmoflmaryflO—fl.19fl,appeunmontheohafl.
mufinmagenufsnorsbontlhomsomasammeue
of that shove “attenuate. lore scan, the 1987 steel no so located on!
Watseehsueemettsnpeuuanmtheemammtmn
Webmaseaseaeppearsuesoaahle. mom-mm»
mmmrmmrmumumnmmmemmmem.
Mmmnmttmeuyhmmueothepoemmvottu
mmfiermmvm.
m,nuammmmpnsunsmamnuum
Mmumm,the1987seonsmstohethesestsevmthae
weseaeeaauybeupaseedeomomtheremesseeumbamabou
Nemataetuoffismtmutepmamenod. It
tummtumummsmmmmmm
Motmufloothemflhionnrandthsttheoowusme
aneslargesauotrelsuvelymununwsnudmofl
data is available for mom study.
mel-dsyetonoflsrohfl-fl,1918,tellenexeontheohefl
nus-Mummy or shoneelnohesomufioo square sues. It
Wetthemyheizhoetehemttmogsnoodsesmmm
ammtfumm-thefennssseenmbasin. Itproduoedthe
”MotnmdumuniummanoodMoheqoanedehe
manuammumionm.mmuunnnmumora

I.
although it has since been exceeded by that or January, 1957. The great
dean‘s eansed.by this flood to industrial cities in Ohio led to the
{creation of the lieei conservancy Dist-riot and the motruotion of a
systee ot.tlood oestrol reservoirs. thus. this store also produced a
(reetfloodoaalearbydreilexeereaandhubeeastudiedindetailby
eecineers ot'theilisei coaserteeer -Distriot so that a large aeount of
asthentie iatoreation on both rainfall and runoff is available.

Il.the applioatioo of the rational runotr’sethod. the percentage
of store reisrsll ehidh runs otr'as flood tloe’aust‘be estiestsd. niece
the laxilne flood'leet occur in the einter season when evaporation and
truepirstioaaremloesndislihelytobspeeededhyrsinaehioheill
leave the mud saturated. the pcoeetege e: wrm that finds its say
leeediately to the streass is likely to be large. lxhibit 26 contains
hydrographs of the principal floods at Chattanooga with sass curves for
each shoeing the onlnlativs rainfall and runoff, ronotf here being store
runoff or the differenoe betseon the recorded flows and an estimated base
flow or ground ester flow. The total runoff for each store is seen to
. be a large preportion of the store rainfall. The following tabulation
times the total rainfall and runoff of lea-gs floods at Chattanooga and the

peroent of the store rainfall ehich appeared so flood runoff:

new mm: .
Influfll Duration Inches Duration Inches Rnnott
'- 13-” (Total) we (Total)

'Cbe-I'Il‘s 1875 14 11.4 82 1.6

tabular. 1517 10 5e5 50 5.8 70
“fr-‘93s 1920 16 6.0 50 3e? 52
MeJue 1931 11 8.0 80 5.8 68
.Ire-‘Pl'e 192. 9 3e]. 80 8.5 7‘

these percentages appear rather low ehen ooapared with those
givee.ie.the following table extracted tron 'Surfaoe'laters of Tennessee,'

57
Bulletin 40 of the Division of Geology, State of Tennessee, by lhrren
s. nag, District lnginecr of the u. 5. Geological Survey:

Wm:- . 122m
him when inf has; unorr Rune
(sq.ai. ) inches acre-ft.) acre-ft.)

Gusherland Barbourville 4.55 258, 000 226,000) 95
Gusherlsnd One. lhlls 2,010 4.9 525, 000 505,000 98
Queberland Burnside 4,890 5.0 ~ 1,500,000 1,190,000 91.5
Gusherland Galina 7,520 4.58 1,780,000 2,515,000 9150
Gusherland Carthage 10,700 5.25 5,000,000 5,544,000 #118
Culberland fleshville 12,800- 5.12 5,480,000 5,840,000 9110
Bookcastle Bookcastle Spa. 746 4.2 162,000 150,000 95
low River low River 512 8.0 155,000 125,000 94.5
5.Fork Cu. lewalsville 1,260 8.08 408,000 586,000 94.8
colline lolinnville 824 7.50 249,000 228,000 91
Obey' Byrdstown 418 5.28 117,000 114,000 97.5
Stones fleyrna 552 5.90 174,000 165,000 95
Cane: fork Rock Island 1,840 8.45 740,000 894,000 95.7
Caney Fork Silver Point 2,100 8.59 958,000 912,000 97.5
leery Harri-an 795 9.00 524,000 507,000 94.8

*iainfall figures fOr Celina, Carthage and lashville do not include
i-inoh of rain larch 28-27 and 1.50 inches of rein larch 50, but runoff!
fros these rains are included in runoff figures.
this tabulation refers to the floods of larch 25-51, 1929, on the Gusher-
land and Tennessee River basins. The store producing these floods
centered over the Gusherland lountains which fore the divide heteeen
these two watersheds. lesions floods of record were produced on cone of
these streaas, but the store did not cover a large enough area on either
the Tennessee or Onshorland Rivers to produce sari-us floods on these
streaes. On the heels of these figures, it would appear that 90 percent
of the rain falling during a store period eight appear as runoff in the
resulting flood period if the store occurred in the winter season and
was preceded by ainor stores which left the ground well saturated.

The working date necessary'fcr computing flood runoff fro-
rainfsll by the rational aethod has been asceablsd, and this data oust
not he applied to the coeputation of the serious flood to be expected

at Chattanooga. The store of’larch 22-27, 1915, was transposed to the

58

Tennessee River basin above Chattanooga, and a hydrograph was computed
by the rational nathod for an scanned runoff of 90 percent. Several
locations of this store over the drainage area were investigated in order
to detersine the one producing the greatest discharge at Chattanooga.
In every case the storn was transposed without changing the direction
of its axis or the shape of the isohyatal lines-that is, the store was
not rotated nor were the isohyetal lines shifted in order to cake the
store fit the shape of the drainage area sore closely. The position
finally chosen is given by the solid isohyetals on Exhibit 27.

Previous studies have showed, however, that the 1915 store
had an average rainfall of only 9 inches over a drainage area equivalent
to that above Chattanooga, while the 1957 store averaged 11.8 inches
. over a sinilar area. The duration of the 1915 storm is generally given
as five days, or about equal to the minimum estimate for the drainage
area above Chattanooga, various longer periods of rainfall say be
chosen for the 1937 store. It is therefore reasonable to suppose that
the 1915 store is not quite the maxinun to be expected over a drainage
area of this size and that it night be preceded or followed by a secondary
store which would bring the total rainfall up to an amount comparable
to that of the 1957 storm. Accordingly, a hypothetical stern of maxi-um
intensity of 2.5 inches was assuned to occur on March 29 and to be so
located that it would contribute to the crest of the flood at Chattanooga.
The ieohyetals of this storm are shown as dashed lines on Exhibit 27.
lith this added rainfall and a runoff of 90 percent, the rational method
resulted in a flood at Chattanooga having a crest discharge of 850,000 cfs,
a total volune of 5,540,000 dayasacond-feot, and a duration of 12 days.

The reduction in crest discharge caused by channel storage

has not been considered up to this point. If this is accused to anount

69

to 14 percent of the crest discharge, the crest is reduced to the
750,000 cfs given by the diagram of maximum Observed Runoff Rates. A
crest reduction of this amount due to channel storage may not be un»
reasonable, although it appears somewhat small, and the resulting dis—
charge is well supported by points on the diagram of observed rates

of runoff. s new hydrograph having this crest discharge was sketched
from the computed hydrograph, the volume and duration being made the
same and the shape similar to the computed hydrograph. Care was
exercised to make the crest of the new hydrograph occur on the falling
side of the computed hydrograph so that the area between the two hydro~
graphs (representing the volume of water held in channel storage) is a
maxi-u: at the time of the crest of this adjusted hydrograph. The
final hydrograph is reproduced on Exhibit 28.

The storm of JanuarylZ-ES, 1957, was also transposed to
several positions over the drainage area above Chattanooga, and the
resulting hydrographs were computed by the rational method with a run-
off factor of 90 percent. The position which results in the greatest
discharge is shown on Exhibit 29. The computed discharge of 950,000 cfs
was considered to be reduced to the previously fixed 750,000 cfs for the
effect of channel storage, and the hydrograph shown on Exhibit 50 was
drawn to have a similar shape and the same volume and duration.
laxilul cod 0 re h

In order to bring together for comparison various possible
flood hydrographs at Chattanooga, a mass-duration curve was prepared
for each and plotted as shown on Exhibit 51, the daily flows being
summed in order of magnitude downward from the crest of each hydrograph.

lass—duration curves were first computed for the floods produced by

60

the transposed storms of 1918 and 1957. It is noted that these curves
are coincident for the first three days, the mass of the 1913 flood
being considerably-smaller than that of the 1957 flood for longer dura-
tions. This appears reasonable inasmuch as the 1915 store was much
shorter than the 1957 store and contained considerably less values when
the entire 1957 stern period is considered.

lass-duration curves were added for the four great floods of
record at Chattanooga. These curves fall far below those of the 1918
and 1957 storms, inasmuch as the crest discharges were much smaller
even though the durations were about the same. The daily flows of the
great floods at Chattanooga were then increased in the ratio of the
. estisated maxi-us expected discharge (750,000 etc) to the actual dis-
charge of each of these floods. This is merely an application of the
basic principle of the unit graph: a storm of given duration over e
drainage area produces a flood hydrogrsph of definite length regardless
of the nagnitude of the store. The added assumption here is that great
floods are likely to be caused by great storms whose durations (duration
of the principal rainfall period) are approximately equal to the time
or concentration of the drainage area. The duration curves of these
enlarged floods all fall below that of the 1957 transposed, the enlarged
1375 and 1886 crossing the 1957 at a duration of 11 days and continuing
to rise sclewhst above it. The mass-duration curve of the flood result-
ing from the 1957 transposed store thus forms an enveloping curve which
contains the mass-duration curves of all the other floods over a ten-day
period surrounding the day of sexinun discharge, the enveloping curve
being exceeded on the 11th day only because base flow was not taken

out of the actual flood hydrographs and so was increased by the same

ratio as the crest flow to give unusually great volumes for long durations.

61

The picture presented by the mass-duration curves is rather
relarkable. The crest portions of the hydrographs produced by the 1915
and 195? stores transposed to the Tennessee River basin are very similar
in shape, although the 1957 hydrograph is broader-that is, the length is
greater at low discharges. The mass-duration curves of the four great
floods of record at Chattanooga fall generally between these two curves
when the daily flows of these floods are increased to bring their crest
discharges up to the estimated maximum discharge of 750,000 cfs. The
less-duration curve of the 1957 storm can be considered as the envelop-
ing curve of all these individual curves.

To summarise, it has been shown that the storm producing the
flood of January, 1987, on the lower Ohio and Hississippi Rivers is the
greatest storm of record that has occurred under conditions such that
its repetition over the Tennessee Valley above Chattanooga at a season
favorable to producing a great flood at Chattanooga is a reasonable
possibility. This stern has been transposed to the drainage area above
Chattanooga in the position which produces the greatest possible flood
at Chattanooga, and the resulting hydrograph has been computed by the
rational method. The crest discharge has been decreased for the effect
of channel storage to 730,000 cfs, the maxi-us discharge indicated for
this drainage area by the diagram of laxinun Observed Runoff Rates, and
a new hydrograph has been estimated which would have the same volume
and duration as the computed hydrograph, a similar shape, and the proper
treatment of channel storage. A similar hydrograph has been prepared
for the 1915 storm transposed to the basin with an added secondary
stern. less-duration curves were prepared for the floods produced by
these transposed stores and for the four greatest floods of record at

Chattanooga with their daily discharges increased in the ratio of the

,‘D

62
maxi-us expected flood to the actual floods. The mass-duration curve.
of the flood produced by the 1957 store transposed forms enemweloping
curve for the other curves. It is thus concluded that the hydrograph
produced by the 1957 store transposed to the valley above Chattanooga
is a reasonable detenaination of the maximu- flood to be expected.under
natural conditions at this point, and Exhibit 52 reproduces this hydro-
graph under the title 'Hydrograph, maximum Assumed Flood, Tennessee

River at Chattanooga, Tennessee.‘'

68
O IWASSEE ER T SSEE AM 8

W

Hiiassee Dam is the second major tributary project to be under-
taken by the Tennessee Valley Authority. It is located on the Hiwaesee
River 75.8 miles above its mouth and about 113 river miles above the
city of Chattanooga. 1 knowledge of the maximum flood to be expected
at this point is of importance in planning the entire project and in
determining the design flood on which the size of the spillway must be
based. The methods which may be used in arriving at a reasonable estimate
of the hydrograph of this flood are similar to those discussed in connec-
tion with the determination of the maximum flood on the Tennessee River
at Chattanooga and so need not be repeated in detail. However, the
drainage area is much smaller, the tapography much more rugged, and the
available rainfall and streamflow data less abundant and probably less
dependable than at Chattanooga, so that the two situations are not
altogether similar.
nggggigtiog of the gagin

The fliwassee River is the first major tributary to enter the
Tennessee River above Chattanooga, and it is therefore important that one
or more storage reservoirs be constructed on this stream to assist in
controlling floods and that these structures be made capable of with-
standing any floods which may be reasonably expected to occur. The
river rises in the mountain ridges along the southern end of the Great
leaky Mountains to flow northwestward to the town of Murphy when; it
is joined by the Hottely and Valley'Rivers to form a fanpshaped drainage
basin. Below lurphy, the river continues to flow through a mountainous
country until it is joined by the Ocoee River, also rising in the

64
lountains and flowing roughly parallel to and west of the Hisassee
River. Below the south of the 0coee River, the aountain ridges become
less frequent and the river flows through a farming country to join the
Thanessee River about 37 miles above Chattanooga.

lhile the loser one—third of the basin is generally rolling
and suitable for farsing, the upper two—thirds is very rugged with sharp
ridges and deep valleys. The river cuts through a number of ridges to
form deep canyons which have been long studied as potential dam sites.
The ridges forming the basin boundary on the south rise to elevations
between 2000 and 5000 feet, while elevations as high as 6000 feet are
reached along the eastern divide between this basin and that of the
Little Tennessee River.

The gradient of the Hisassee River is generally steep. The
slope averages slightly over a foot peraile over the lowest third of
its length, reaches a narinun of 600 feet in 10 miles where the river
passes through the more rugged country, and averages over 15 feet per
sile over 60 silos of its length.

The average annual rainfall over the Bivassee River basin is
about 10 percent greater than that over the Tennessee River basin above
Chattanooga, about 57 inches as conpared with 52 inches. The average
runoff is slightly over 50 percent of the rainfall.

The site of Hiwassee Dam is in the mountainous portion of
the basin there the stress profile is steep and the valley narrow. The
selection of a site far enough downstream to control a sizesble drainage
area and yet not too far downstream to be out of the mountainous section
with its narrow and steep valleys and its good den sites places this

das about 20 miles below Murphy and therefore immediately below the

65
fan-shaped drainage area for-ed by the Junction of the Valley and Hottely
Rivers with the Hiwassee River. The drainage area at this point is 977
square silee.

W

A nulber of streaeflow stations have been Operated on the
Hiwessee River and its principal tributaries by the U. 5. Geological
survey. Of particular importance are the gages on the Hiwassee River in
the vicinity of Reliance. The drainage area at this point is 1180 square
silos, or only 12 percent greater than the drainage area at the dam site.
The first gage was installed at this point in 1900, end stage records
are continuous to date, although the gage has been moved several times
without obtaining a sufficient number of discharge neeeurenents at each
location to adequately determine a rating curve and without obtaining
overlapping records of sufficient length to deter-ins the stage relation
existing between gages at the different locations. A gage was established
on the Hiwessee River at lurphy in 1897 and records have been kept to
date. The drainage area at this point is only 419 square miles, since
the gage is above both the Valley and Nottely Rivera, but the records
are of particular value in studying the backwater protection.which sust
be provided through the town of lurphy.

A third gage of isportance was established at Charleston in
1898 on the lower portion of the river below the south of the Ocoee
River, with a drainage area of 2296 square eiles. The period of record
is not complete, but stages from e staff gage read by the U. 3. leather
Bureau in this vicinity are useful in filling out the record. Flood

discharges detersined fro: the recorded stages must be used with caution,

66
however, since Charleston is below the steep portion of the river and
stages at this point are affected by backwater free the Tennessee River.

Prior to the establisheent of a large nulbsr of rainfall
stations in the Hivessee River basin by the Tennessee Valley Authority,
the station at Iurphy was the only one within the drainage area above
the due site. Only six stations were operated at any one time within
or close enough to the basin to be of such assistance in estimating
store rainfall over the area. Considering the great variation in eleve-
tion throughout the basin and the effect of the mountain ridges upon
soisture-laden winds, these stations are far from satisfactory in any
store studies where it is necessary to determine average rainfall over
the drainage area. The neny additional gages installed by the Tennessee
Valley Authority are a great help, but the records from these gages are
too short at this time to be of such assistance.
MW

Although empirical flood flow equations were found to be of
little assistance in estimating the marines flood to be expected on the
Tennessee River at Chattanooga, they should be more successful on the
Hiwaseee River since the drainage area involved is more within the lisits
of the data from which these equations were originally develOped. The
sountainous drainage area above Hivaesee Dan Site, however, might be
expected to produce floods somewhat greater than those given by scat
aspirical equations.

Suichling's equations derived from the lohawk River give
Rare floods, o = :—2§‘%$% / 7.4 - 100,000 cfs
Occasional floods, Q . {'WU / 20 a 57,000 cfe

The lnrphy'equation derived free northeastern united States

6?

gives a a :5 7:30 ,l 15 a 54,000 cfs.

The layers equation for extreme floods as sodified by Jarvis
gives :2 = 10,000 a?! a 510,000 cfs.

The Fuller equation with 'C' estiseted at 70 and '1" taken as
1000 years gives Q I C ‘0.8 (l ,1 0.8 loglo T) I 58,000 cfs.

The general width forum of C. R. Pettis with a l-day rain-
fall of 7 inches and an average basin width of 14 miles gives a probable
loo—year flood of c = 430 r w1o25 s: 91,000 cfs.
W

Ctegeenddischargereeordecoveringaboutmyeersareavail-
able en the livassee liver at Inrphy, Reliance, and Charleston. The
record at Reliance is of particular interest since this gage is nearest
to thedee site inbotheileageenddrainage area.

bequency curves were prepared by the Corps of Ingineers for
the gages at Iurphy and Reliance using the Goodrich sethod, end the
results were published in a. D. 528 as previously sentioned. The
SOC-year flood at each of these gages is plotted along with corresponding
floods at other gages within the Tennessee River basin on Exhibit 15,
and an average line is drawn to represent the relation between drainage
area and the ceeputed SOC-year flood. This line passes directly through
the point plotted for Reliance and cones very close to the point for
lurphy but falls considerably below the points representing long periods
of record at Knoxville, Chattanooga, and florenee.

In the frequency studies described in connection with the
deter-instion of the latitu- flood to be expected at Chattanooga,

20 and 60-year periods selected free the 70-year record of floods at
this gage resulted in 500 and loco-year floods of considerably differut

1‘1“.

68
lepitndes. The e0-year periods of record on the Kiwassce River can
not be expected to produce frequency curves which sight be extrapolated
to 500 years with any degree of certainty, although they night he very
usefulinstudyingthe-allerflcodsthatarelikelytooccurevery
few years. Further, an alaninaticn of the store rainfall in the Kiwassee
River basin indicates that this area has never been subjected to stores
as severe as those which have hit surrounding areas in the Tennessee River
basinandwhichsightbesnpected toocourover this drainagebaain.
Thus, the period of available records is not a fair sasple of conditions
beeauseitisshortanddoesnotcontainstoresasgreatashaveoccurred
over adjacent drainage areas.

The uncertainty of the nagnitude of the crest discharges of
thehiafloods ofrecordatbothlelianoeandcharlestonnnst alsobe
considered in attupting frequency conputations based on available
knowledge of past floods. a discharge of 56,000 cfs has been estisated
for the flood‘of April 2, 1920, at Reliance, the highest flood of record
at this gage, while 65,200 cfs is reported for the flood of love-bar, 1908.
lothfiguresaresubjecttocorrectioninthefutureassoredisohsrge
censure-eats are nade at high flows. The flashy nature of the stres-
nakes flood peaks very short so that it is quite difficult for an
uginesrtoreachthegagingpointandaakeadischargeaeasureaentwhile
a flood is near its crest.

W

n. 'unit mpa- and m "a... Manama. have been
described as one of the nethcds available to the hydraulician for convert-
13g rainfall into runoff, and cenpntation of the flood produced when a
lazin- store is transposed over the drainage area is one of the
possible applications of this nethod.

89

is previously explained, it is often difficult to find floods
which were produced by stores of unifm intensity and duration throughout
thewholedreinagcarcafcrnseinpreparingaunitgraph. Thisproved
to be the proble- in studying the Eiwassec River basin. Rainfall records
prior to the establish-fit of the Tennessee Valley Authority gages
were fed to be of little value, since only one station was located
within the basin and the great variation in elevations would not per-it
statim sue distance outside the basin to be used in ascertaining the
nnifcrsity of the intensity and duration of the rainfall over the
drainage area. In the short periods of record available since Tennessee
Valley Authority rain gages were established, only a few floods of any
sine have occurred, and the rainfall producing these floods did not
have the required unifornity. consequently, the unit graphs which were
deter-iced fron these data did not have a great deal of sinilarity;
variatioasinpeakdischsrgersngedashighaswpercant.

An average unit graph constructed froc these individual nit
graphs could hardly be imrcased the anount necessary to obtain the
sexin- flood, nor could the unit graph having the sexism crest discharge
be increased in this nenncr without obtaining results which night he
enpcctedtobeinerrorbysonethinglikeflorSOperccntbecauce ofthe
inherut difficulty of obtaining a reliable unit graph.

The hydrographs of these floods and the resulting unit graphs
were useful, however, in checking the shape of the hydrographs obtained
by the rational nethod and in drawing the final hydrograph of the

 

The cannu- recorded runoff at all gages in the liwaesee liver
basin is chm on Exhibit as. The short pcriods of record at these

7.
gages and the apparent lack of any great stores over the basin within
this period cake it advisable to consider the serious rates of runoff
that have occurred in other basins having sinilar flood producing
characteristics. is previously discussed, Exhibit 7 represents recorded
rates of runoff throughout eastern United States, with the exception of
areas along the southeastern and southern coasts share heavy runoff is
produced by tropical hurricanes. The hivassee River basin is well
shielded by high contain ridges so that these tropical stores are not
likelytoprodueeheavyreinfall overlargeareaswithinthebasin.
Thelowparoentageofrainfallwhiahappearsasrunoffdtn‘ingthesuneer
scnths when stores of this type occur will further tend to produce loe
runoff rates.

Theenvelopinglineshownonthis diagranwasrecoanudedby
the 0onsulting Board for general application to drainage areas in the
Tennessee River basin of sore than 600 square niles. is previously
noted, however, the Board recognised that in dealing with particular
drainage areas 'each problee should be Med in detail with the
exercise of Judglent in the light of the local conditions' and added
further: 'In sole cases it nay be necessary to natarially increase or
decreaeethevaluasecdcterwincdinordartoallowforspeciallocal
conditions.‘ _

. It is notedonthediagrsn oflannunhunoffhatcs that the
stain of larch 22-28, ‘1929, over the Tennessee and Ounberland basins
resulted in a point representing a discharge of 198 ofs per square ails
frce a drainage area of 800 square nilcs, this point lying sonewhat above
thesnveloping line. Thisfloodoocurredcnthelaoryniver atnarrinan,
a drainage basin sinilar in shape and topography to that of the

liwacsee River basin above livassce Den Site and not far distant fro-

this area. This rate of runoff is based largely upon discharge
neasureeent’aade under difficult conditions and is therefore subject

te cone doubt. It nay'be either too high or too low. However, when

this high rate ofheunbff is considered alongdeitb the fact that the
liwassee liver'basin has steep slopes, no surfeee storage, a fan-shaped
drainage pattern, and a soil and rock cover which is highly inparvious

te water, it is reasonable to expect this basin to produce floods greater
than those occurring on cost basins of this siae in eastern.0nited Btatec,
and the enveloping curve on Inhibit 7 should be raised to include every
point rather than Just the great naJ ority of point}. in enveloping

line having the equation 0 : /£QQQ. rather than.0 : .gggp, is considered
to represent a reasonable ectiaate of the sari-us di‘oh:rge to be expected
free drainage .1»... in the upper portion of the liwaesee liver basin.

This gives a runoff rate of 10! efs per square ails at the den site of a

discharge of 100,000 cfs.

 

. The so-called 'rational' nethod of computing flood flow fron
rainfall nay be applied tothe Hiwassee River basin in much the cane way
that it was used in detereining the naxinun flood on the Tennessee River
at Chattanooga. The general principles of this method were discussed
previously and need not be repeated.

Tine cones were computed at 6—hour intervals over the Tennessee
River basin above Chattanooga, and these nones on the Hiwassee River may
also be used in studying floods above the Hiwassee Dan Site. is shown
on IXhibit 15, there are only three time zones above the den sits,
indicating a period of concentration of between 12 and 18 hours. The

6-hour cones are rather large for this small drainage area, but smaller

cones would probably give an appearance of refinement which could not

72
be attained with the available data.

0f the three types of storms previously discussed as occurring
within the Tennessee River basin, cloudbursts result in intense precipi-
tation over only a fee square miles and generally spread over less than
a hundred square miles so that they need not be considered as possible
producers of a maximum flood at the dam site with its drainage area of
nearly a thousand square miles. Both cyclonic storms and West Indian
hurricanes lust be considered.

The annual and seasonal distribution of floods at the Reliance
gage of the U. 8. Geological Survey is pictured on Exhibit 54. The
record is complete from 1901 to 1057, with the exception of the period
from 1914 through 1918 when the gage was located at Apalachis, 16 miles
upstream, and stages could not be related to the present gage with
reasonable accuracy for comparison with other floods. The seasonal

distribution of floods is seen to be similar to that of the Tennessee
River at Chattanooga as pictured on Exhibit 4, although the flood season
appears to be somewhat longer, starting a little earlier in the full and
continuing s little later in the spring. It is noted also that the maximum
flood sheen occurred on November 19, or very early in the flood season,
while the second highest flood occurred near the end of the flood season,
on April 2. There are only a fee summer floods and none of any magnitude.
This diagram suggests, therefore, that winter storms of the cyclonic type
are the principal producers of floods at Reliance Just as at Chattanooga
but does not prove that great floods night not be produced by summer
stores of the West Indian hurricane type.

in examination of the tabulation of the maximum floods of

record at the various gaging stations in the Hiwsssee River basin as

78

given on llhibit 55 shows that the West Indian hurricane of July 6-10,
1916, produced the sari-us flood of record on the 0coee River at both Eat
and Plrksville, even though all the canine on the Hiwassee River itself
were produced by other stores. The path followed by this hurricane is
pictured on Exhibit 19. Here it is noted that this storm had an intensity
of about 15 inches just before it entered the Tennessee River basin,
dropping to only 6 inches in Tennessee. It was explained previously
that the lest Indian hurricane of July 14-16, 1916, also resulted in the
greatest rainfall Just outside the Tennessee River basin, although the
rain which fell upon the upper portion of the French Broad River was
great enough to produce the maxi-um flood of record at many points on
this stress.

in examination of the contours on Exhibit 1 discloses a gap
in the divide along the southern edge of the 0coee River basin; the
elevation drops below 2000 feet for a short distance along this divide,
whereas the divide generally ranges from 5000 to 5000 feet above sea
level. There is some possibility that a West Indian hurricane night nove
up through this gap to cause excessive precipitation in the valley, but
the narrowness of the gap and the high and rugged ridges surrounding it
would doubtless prevent such a stern from covering an area the sine of
that above the den site with rainfall of sufficient intensity and
duraticn to produce the naxiwul flood at this point.

It must also be considered that'Iest Indian hurricanes occur
in the suller and early fall uonthe when the percentage of rainfall
that appears as runoff is low, even in the case of intense storms of
short duration. To produce the ease runoff, a West Indian hurricane
lust therefore have an intensity aleost twice as great as that of a

winter store, and it is not reasonable to expect a hurricane of this

74
intensity to enter the Hiwassee River basin and to cover an area the
else of that above the den site.

Btorss of the cyclonic type are the remaining class to be con-
sidered as being capable of producing the naxilun flood at the den site.
As shown on Exhibit 20, storns of this type seen to follow up the basins
of the Ohio, Gueberland, and Tennessee Rivers, and it is reasonable to
expect that one say center over alsost any portion of the Tennessee
River basin.

To be applicable to the present case, such a storm must have
a duration of less than one day, as the ties of concentration for this
drainage area is between 12 and 18 hours. The store of March 22-25, 1929,
previously referred to, which centered over the divide between the
Culberland and Tennessee River basins is a good example. This store
produced the sari-us flood of record on the Enory River, a tributary of
‘the Tennessee River system, and on various tributaries of the Culber-
land River system, causing property dosage approaching $5,000,000 and
drowning at least 22 persons.

A considerable amount of date was collected on this flood and
the stars which produced it by l. in. ring of the U. 3. Geological Survey.
lo recording rain gages were in operation within the store area, but
issediate field investigations were undertaken and the observations of
the regular leather Bureau stations supple-ented by the measure-ant of
rain collected in cans and tanks. the store began on the earning of
larch 22 and continued until the next warning, but local observers
reported that seat of the rain fell within 18 hours and about half of it
fell in sosething like two hours.

The isohyetals of this stern are drawn on Exhibit 25, and the

computed tine-area-depth curve is shown on Exhibit 25. It is noted that

75
the tise—ares-depth curve for this store falls very close to curves for
the storss of lay 51-June l, 1889, and love-her 5-4, 1927, and is exceeded
only by that of October 5-4, 1869, the naxisus rainfall being about 11
inches while nearly 9 inches was averaged over an area of a thousand
sQuare siles.

The stars of larch 22-25, 1929, appears to be a reasonable
stors to transpose to the Hisassee River basin to produce the saxinua
flood. The tine—erea-depth curves show it to have been exceeded only by
the stars of October 5~4, 1869, which occurred outside of the flood
season. The flood discharges caused by this store plot very high on the
diagras of eaxisus runoff rates. The store had about the proper duration,
and it occurred on a similar drainage area located very close to the
Hiwaseee River basin. Further, a considerable amount of reliable data
has been collected on this store and is available for study and use.

Since no recording rain gages were in operation within the
area covered by this store, sone assusption has to be made as to the time
of occurrence of the rainfall before the stors can be applied to the
Hiwassee River basin. Based on the available information regarding the
stern, it is assumed that all of the rain fell in a 24-hour period:

20 percent within the first six hours, 40 percent in the second six,
50 percent in the third, and the reeaining 10 percent in the fourth
sixrhour period.

The percentage of rainfall which is likely to appear as runoff
during a major flood on the Hiwassee River cannot be estimated from
'eveiiebie records of rainfall and streanflow in this basin. Rainfall
stations are too scattered prior to the establish-eat of a large number

' of stations by the Tennessee Valley Authority to give an accurate

78
picture of the average rainfall over the drainage area, and a sufficient
nuiber of floods of any sise have not occurred since these stations were
established. The investigations node in connection with the application
of the rational runoff nethod to the Tennessee River at Chattanooga
indicated that an average runoff of 90-percent during a flood period is
reasonable for that area, and the sane figure appears reasonable for the
Biwaseee River basin above the dam site.

The store of larch 22—25, 1929, was transposed to the basin
of the Hiwassee River above the den site with the area of maxi-us rain—
fall located ilnediately above the den site, care being taken not to
rotate the axis of the store in transposing it to this location. The
rain was all assueed to occur in a 24~hour period with the distribution
previously explained, a runoff factor of 90 percent being applied in
computing the flow at the do. site for each 6—hour period. The resulting
hydrograph does not contain the effect of channel storage in reducing and
broadening flood peaks but gives a reasonable picture of the hydrograph
that night he produced by a store like that of 1929 if it were to occur
over the Hiwassee River basin in the position assuled.
W

The crest discharge of the eaxinue flood hydrograph has been
estieated at 188,000 cfs free a study of the naxiwue rates of runoff
observed on streans'haying similar flood producing characteristics, but
it reeains to estimate the volume and duration of this flood and to
sketch a hydrograph having a reasonable shape.

The hydrograph coeputed by the rational nethod for the storm
of larch 22-23, 1929, transposed to the area above Hiwassee Dan Site
nay have its crest discharge reduced by channel storage to the 188,000 cfs

previously detereined without change in duration or total flood voluee.

77
The resulting hydrogreph eppoere generally similar in duration and ahape
to many of the unit graphs ooeputod for the Hieaaaee River at Reliance.
This hydrograph, therefore, has a crest discharge determined from maximu-
fleode which have occurred over eieilar drainage areas on other rivers
in eastern United statee, e duration equal to that produced by tranc-
poeing the 1929 ltorl avor the drainage area and a shape similar to the
unit graphe at Reliance and the coeputed hydrograph of the 1929 store

transpoeed, and eo represents a reasonable estimate of the maxi-u: flood.

78

mgggflggzon 01" DESIGN 21,922

m
In discussing the methods commonly used to estimate necessary

spillsay capacity, the problelsas divided into four parts: Detersination
of (l) sainUI flood hydrograph, (2) project flood, (5) design flood, and
(4) planning the spillsay.

The various methods commonly employed in estimating the dis-
charge hydrograph of the maxi-us flood reasonably to be expected at the
site under the conditions existing prior to construction of the project
or any other eater control project within the contributing drainage area
sore reviesed in a general say and then illustrated by reference to the
Tennessee River at Chattanooga and the Hissssee River at the Hisaseee
Den Bite, the first representing a drainage area of 21,400 square miles,
shile the second represents less than 1000 square miles.

The "project flood" differs from the "maximum flood' in that
it is the greatest flood reasonably to be eXpected at the site under the
conditions existing prior to construction of the project and thus in—
cludes the effect of any upstream regulation, while the maximum flood
is based on natural conditions Iithout any upstream regulation. The
project flood say be determined from the maxinun flood by computing the
effect of upstream storage, or it say be the result of a store centering
over the uncontrolled drainage area between the upstream projects and
the point under study. Both possibilities must be investigated. The
rational method of estimating runoff frOm rainfall say be used with
hydrographs computed at each of the upstream projects and these routed
through the various reservoirs and finally down the river channels to

the site being studied.

79

the 'delign flood' results free routing the project flood through
the preposed reservoir, consideration being given to the effects of
natural and controlled storagee within the reservoir upon the original
flood hydrograph. Binoe the natural channel storage existing within the
length of river covered by the reservoir is displaced by the reservoir
storage, the effect of this natural storage oust be taken out of the
bydrozraph of the project flood or it will be considered twice in the
routing ooeputaticns. This requires another routing under natural
channel conditions to detarsine the hydrogreph produced by the project
flood at the upper end of the reservoir under the natural channel condi-
tions existing prior to construction of the project. In case the next
upstreas project is immediately above the proposed project, its outflow
hydrogrsph will be the required inflow hydrograph, and in cans the natural
channel storage is snail, the project flood say be taken as the infloe
hydrograph rithout serious error. the hydrograph of the local inflow
which is discharged directly into the reservoir is to be added to the
inflow hydrocrsph to obtain the total inflow. This hydrograph is then
equal in voluse to that of the project flood but has a shorter duration
and higher crest discharge since the flattening effect of the displaced
channel storage has been elininated. It is this total inflow hydrogrsph
which oust be routed through the reservoir to obtain the design flood.

In routing the total inflow hydrozraph through the reservoir to
obtain the design flood, the scat severe conditions that eight be
reasonably expected are assused to exist. The project flood is generally
seas-ed to occur sith the reservoir at the highest level at which it can
be saintained eith all the gates closed. The gates reeain closed until

the store has struck and then are opened, at first only sufficiently

to hold the pool elevation constant by asking outflow equal to inflow

80
and then fully as the seriousness of the flood becomes apparent, in order
to lower the pool and provide additional capacity for storing at the time
of the flood crest. Sons allowance should be made, however, for difficulty
end delay in gate operations. It is often assumed that the power plant
is closed down so that no flow is discharged through the turbines and
that several, or all, of the sluices have become clogged with debris
or the gates stuck so that they cannot assist in discharging the flood
esters. If there are e large nunber of gates to be opened, it may also
be reasonable to assume that some of these gates are jammed with debris
and cdnnot be Operated. I
The routing procedure generally followed in determining the
effect of s reservoir upon s flood or in computing the design flood
free the total inflow hydrograph is s step process using inflows at equal
intervals of tins and ccnputing outflows at similar intervals, the
difference between inflow and outflow over the ties interval considered
being equal to change in reservoir storage. If inflows at the beginning
and end of s tine interval 't' are denoted by '11” and '12' with 'Qll
end '02. representing the corresponding rates of outflow, and '81“ end

'82. representing the reservoir storeges, the equality'msy be written:

W-(Q1/Qz)t-82-sl
2 ...__..___.........

2
If the tine interval is taken es 24 hours and the flows expressed in
cubic feet per second, the terms expressing everage inflow and sversge
outflow will be in day-second-feet, or since the flows are divided by 2,
each tern represents acre feet. If the storages are also expressed in
sore feet, the equality reduces to 11‘/ 12 - Q1 - Q2 = 82 - d1. '
In the routing process, the inflows are always known while the storage

and outflow at the beginning of each tine interval are obtained from the

81
omtotiono for tho prooooding intorwol, ond thooo waluoo oro “and
for tho bocinnin. of tho firot intornl—thot in, tho gotoo oro umlly
Wehoodoothotthoroilnooutflowotthoboginnin‘oftho
flood poriod, ond tho pool io oonoidorod at tho top of tho gatoo to fix
tho otorogo. magmammmymM-mmoqmuty. lhon
thooo unknom oro tronopoood to tho right oido of tho oquotion, thou-o
roonltoc 11/121181-013Q2/82.

tho otorogoo horo omidorod not ho tho totol mount of
“tort-porcih'ota'odhthorooorwirot thoinotontoftiooboin;
moidorod. an o largo riwor whoro tho rooorvoir io loos, tho olopoo
my not, ond tho flow through tho rooorwoir rolotiwoly lorgo, tho
otorocouodinthoooloulotiono emotbothotmdorolowolwotor
mfooo but not ho thot ondor o hookwotor our" for tho hoodwotor ond
flow oooditiooo oxiotiog ot tho tioo. In tho oooo of o tributory
proJoot, tho otoopor olopo ond lulla- flow prodnoo o ooollor nomt of
bookwotoromoohortordiotouoondthononlyinthoupporondoftho
roomoir whoa-o tho pool ooworo tho rim ohohnol only, thuo roaming
hookntor otorozo to on ooount which con gonorolly ho moglootod.

If bookwotor otorogo oon bo noglootod, both outflow ond otorogo
hooooo o function of dioohorgo whoa tho opillwoy xotoo oro oponod, ond
tho right hood oido of tho oqootion, or tho 'outflow-otorago footor,‘
no: ho plottod ogoinot hoodwotor olovotion or ogoinot oithor outflow or
otorogo. Such a :ourwo ooobloo both outflow and otoroco to ho rood
dirootly oftor 'thoir no hoo boon mm by «an. togothor tho
hon quontitioo on tho loft oido of tho oquotion. Mo givoo o diroot
solution of tho otorogo oquotion ond ponito outflowo to ho ooopotod by
oooooooiwo otopo without tho out-ond-try ooopntotiono that would othorwioo

82
ho noooooory to bolonoo tho difforonoo hotwoon inflow ond outflow oaoinot
ohonco in otorogo.

Ifhoohwotorotoruoiotoolorgotoboooclootod.nuorono
booklotor mo not to o-putod for miono hoodwotoo' olovotiono ond
woriouo inflowo o‘ ontflowo. ltorogo thou boooooo o function of tho
throo mun-u hoodwotor olowotion, inflow, and outflow. If tho
rooorwoir‘ io not too long, boohwotor otorogo will ho dopondont lorgoly
upon hoodwotor olowotion on! inflow, ond tho offoot of outflow no: ho
noclootod, thoroby olioinotinx ono of tho voriohloo ond ponitting
ontflow-otorozo to ho plottod. ogoinot outflow in o fooily of onrm,
oooh mo rmooutin: o difforont inflow. 'lhio ogoin 31on o diroot
oolntiou of tho oqnotion of otorogo ond poroito outflowo to ho oooputod
in onoooooiwo otopo.

lfontflowoonnotbonoclootodinooopntingboohwotor
otoroco. tho prohla boooooo ono of plotting four woriobloo, ond
odditionol fuilioo of onrwoo out ho proporod. In tho oooo of wory
long rooorwoiro, otoo-ogoo booooo dopondont upon flowo and olowotiono ot
voriono pointo olong tho longth of tho rooorwoir ond o out-onddtry prooooo
out ho rooowtod to in orriwinc ot ohongooin otoo'ogo.

Tho otoroco oqnotionooybohondlodbyoonyothoroomodo in
rootinxfloodothroughorooomir,hntthooothodoorobooioollytho
oolo old tho rooolto xonorolly idontiool.

mmmmmmbyapwmmmsmm
to tho rooorwoir nndor tho ooot oowoo-o condition to ho moon-bl:
W boo boon torIod tho 'dooign flood.“ If tho olovotion to which
tho pool riooo in thooo oooputotiono mood- thot prowiouoly fixod u
tho hixhoot ollowohlo pool olowotion, tho dioonoiono of tho opillwoy

BB
nothinoroooodbyoithorloworingtho orootor inoroooing ito lonzth.
low rooting olrwoo not ho proporod for tho oodifiod opillwoy, ond tho
prooooo loot bo ropootod until tho ooot ootiofootory and ooonooiool
oolhinotion of opillwoy noun ond goto oioo io found whioh will oofoly
pooo tho dooiu flood.

Forthor opillwoy invootigotiono onot ho oodo in tho fourth
otopofthopoooooifthorooorwoiriotohoooodforfloodoontrol
ood tho otorogo oopooity io not onffioiontly groot to rotoin tho ntiro
flowofo oinclogrootfloodwiththohoodwotorot thohoginninzoftho
flood ot tho olowotioo ot which tho projoot will nor-oily ho oporotod
during tho flood ooooon. ‘
WW

tho W Of mm flood ,roooonoflly,to,bo 'OIPOGtfid
ot tho Iliwooooo no: lito hoo boon dooidod nomad it'rohoino to dotoooioo
tho projoot flood and tho dooia flood boforo tinny fixing tho odoo
of tho opdllwoy.

Atthoproooottioothorooroooworolhydro—olootrio
dmlopontowithinthodroinogoorooohowotholiwooooonooaito, but
thooo dowolopoooto oro oll oooll ond thoir otorogo oopooitioo nogligihlo
widow flood condition. Shooo would, thoroforo, howo littlo offoot
upon tho Ion-m flood hydropoph ot liwooooo D on alto, ond tho projoot
floodio idontioolwiththoooxiouoflood.

rhonotnrolohonnoloftholiwoooooliwor ioworynorrowond
otoopwithprootioollyoofloodploinolongogrootportiooofito
1““. Manual: in tho mtoinono notion of tho wolloy whoro tho
rim pooooo through o. oorioo of ridgoo. 'fhio io oquolly trno of tho
portiooofthoriworwhiohwillhofloododhythopropooodfliwooooonoo

84
oothotthomtofohonnolotorozowhiohwinbodioplooodbytho
noorvoir io vory loll. If only flot pool otorogo io oonoidorod in tho
flood routing oooputotiono, tho noglootod otorogo inorooonto hotwoon
flotpoolondthohookwotoroorwooproduoodbythoflowoooyhooonoidorod
to offoot tho inorooonto of dioploood ohmol otorogo and tho inflow
hydrou-oph will ho tho oooo oo tho hydrogroph of tho proJoot flood.

It io propoood thot flood flowo ot niwooooo Do- to pooood
hyoopillwoyhowinxooloorln‘thofzwfootondhyfour oluiooo
howing o totol oopooity of ohout 20,000 ofo Nor norool hoods. tho
opillwoy oroot io propoood to ho ot olowotion 1508.8 ond tho top of
gotoo ot olovotion 1526.5, tho pool boing hold .1th holow tho top of
thontooduringthoouooorooooonbutouohlowordurinxthowintor
floodooooon. fhopoololowotionionottooxooodlminordorto
prowont uooooiwo upotrooo dougo. fhooo opillwoy dioonoiono provido o
dioohorgo oopooity of 130,000 ofo, or o total oopooity for opillwoy old
oluiooo of ohout 150,000 ofo.

Aooriooofroutingourwooworowoporodforthoooondothor
propoood opillwoy ond oluioo ooobinotiono, ond tho projoot flood woo
routod through tho rooorvoir Indor woriouo ooouood oonditiono to dotor-
oino tho noxiou- dioohorgo ond tho pool olowotion rooohod by tho dooign
flood. looplo routing oorwoo oro shown on Inhibit 80 for o throo-hour
tioo intornl. i lonxor timo intorwol would not furnioh ouffioiont
pointo to dofino tho dooign flood hydrogroph with ouffioiont ooourooy.
In thio ding-on, tho ohoioooo roprooonto rooorwoir outflow in tono of
moo doy-oooond-foot por 5-hour poriod of tioo, whilo tho ordinotoo oro
outflow-otorogo footor, or tho on. of outflow and otorogo. in odditionol
our" in oddod to ohow tho rolotion of otorogo to rooorwoir olowotion.

iopo-owiouolydioouoood, thoooxionfloodio oonoidorodto

85
rooultfruowintorotornofthooyolmwpo,ondthorooorwoirio
tohoholdotorothorlowolowotionduringthowintorooooontoprowido
otoroeo oopooity for oootrolling floodo of thin typo. If tho oluioo

'goltoooroollooounodtoboolooodorthooluioooolowd'flhdibflls

finish on 1510 (8.! foot ohowo tho top of tho opillwoy) ot tho tino of

‘.
, ‘ f
i y I

ooourronoo of tho projoot flood ond otill onoblo tho opillwoy to pooo

fl
,1
I

rioutin oonpntotiono indiooto thot tho rooorwoir olowotion night ho oo

' ' thio flood without onooodinz tho noxinu poo-niooihlo olowotion of 1582.

ainoo it io propoood to drow tho pool down to on olovotion of about 1415
by tho hozinning of tho flood ooooon, tho otoruo hotwoon thio olowotion
ond olovotion 1510 noy ho oonoidorod oo o footor of oofoty.

If tho projoot flood ohould ooour (duo to oouo pooulior ond
unforooon ooohinotion of oirounotonooo) with tho pool at tho highoot
olowotion ot which it oon ho nointoinod, olowotion 1520, routing
oooputotiono ohow that by oponing oll opillwoy ond oluioo cotoo thio flood
could ho oofoly pooood with o dioohargo not oxoooding 130,000 ofo ond
o hoodIotor olovotim not mooding 1580.6. lith o roooonobly rolioblo
forooooting oyoton, tho opillwoy gotoo night ho oporotod to roduoo thio
oroot dioohorgo oooowhot, by holding tho outflow to 110,000 ofo without
ollowin; tho rooorwoir to rioo obowo olowotion 1532.

Q m propoood opillwoy hoo boon found to ho ouffioiontly

lor‘go to pone tho dioohorgoo rooulting fron tho projoot flood, but nony

,Mtionol otudioo not to nodo in oonnootion with port 4 of tho

invootigotion: running tho opillwoy. fhooo studioo nuot noko oortoin

. thot tho propoood opillwoy ond oluiooo oro tho propor oonhinotion to

provido tho pootoot poooiblo flood oontrol bonofito (ot roooonoblo

' out) fron tho ovoilohlo otorogo oopooity of tho rooorwoir. rho oon—

binotion of olniooo ond opillwoy not ponit tho rooorwoir to to drown

00
doonropidlyuponthoopprooohofoaonorolfloodoworthodroinago
oroo ohowo tho pointo to ho protootod ond botwoon twin floodo that night
occur in quick ouoooooion. Buoh otudioo prooont nony now prohlono in
tho ohoioo of twin floodo, tho poooihlo opooing of thooo floodo, ond tho
mood ooh-on of tho flood control oporotion of thio rooorwoir ond
othor roomoiro in tho oyoton for tho flood control honofito to loool
orou and to tho Ohio and liooiooippi limo. Ouch prohlono oro hoyond
tho ooopo of thio dioouooion.

B7

 

tho nothodo ouonly uood in ootinoting tho noxinun flood
dioohorgoforwhioh tho opillwoyofowotor control otruoturo nuotho
dooignod nod tho oppliootion of thooo nothodo to pointo within tho
fonnoooooliworbooinoorwotoduonotroto thodongorofohootyond
iuonploto inwootigotion of out: o prohlon ond point out tho nood for
odditionol hydrologio doto ond inprovod nothodo of onolyning ouch doto.

hginooro mogod in hydroulio invootigotiono oro gaorolly
opprooiotivo of tho innonoo onoxmt of donogo thot oon rooult fron wotor
whiohhooohoowoyfroothooontroloudorwhiohithooboonplooodond
ogroothotonroooonohloproooutiononootboonoroioodtoprowontoah
no mo. tho prooont doy tondonoy tooordo tho motruotion of
rooorwiro of oll oiooo on o noono of orooting Jobo for tho unonployod,
howowow, in opt to rooult in hooty phoning of thooo otruotm-oo in on
ottupt to prowido i-odioto .ploynont ond in tho «who. of
dooigning aginooro on tho booio of politiool affiliation rothor thou
utility ond oxporiouoo. rho otodioo and dioouooiono horo poooontod
Wto that noooooory opillooy oopooity oonnot ho proporly ootinotod
fron o oinglo nothod of otudy nor from tho oppliootion of hondhook
oquotiono but out rooult froo tho oxporiooood Judgnont of o ooopotont
nginoorwhohoonodoothoroughotudyofthohydrologioolondutoorologiool
ohorootoriotioo of tho portioulor droinogo booin.

loinfollondotroonflowrooordoorothofnndooontoldotoof
thohydroulioionongogodinwotoroontrol otudioo, ondolthoughthonood
for o grootor quontity ond hottor quolity of thooo doto howo boon
W by now inwootigotoro, it oon boor ropotition. Iony odditionol
roin gogoo woo-o inotollod by tho fonnooooo Volloy Authority throughout tho

88

Tonnoooeo River basin as tho existing gages were too widely separated
to provide a true picture of store rainfall, particularly in a nountain-
ouo area where elevation and exposure influence rainfall to a high degree.
In a siniler way, our knowledge of stern rainfall inoother drainage
basins nay be increased by tho installation of additional gages in those
areas where gages are scarce, particularly if tho topography is such as
to produce a considerable variation in rainfall over short distances.
intonatio rain gages are particularly useful as they may be installed in
isolated areas with only occasional inspections. in accurate record of
rainfall intensities over short periods can thus be obtained as well
as tho tines of beginning and ending of rainfall periods.

lany additional recording streen gages have boon installed
by the U. 8. Geological Survey in recent years, and an ever greater
nunber would be.useful. It is inportant, however, that each gage instal-
lation be so arranged that stages will be recorded by the gage over the
conplote range in water surface elevation that nay be reasonably expected.
It in oonetinoo impractical to build the float well of a recording gage
of sufficient height to allow the gage to record the naxinun flood to
be expected, but every effort should be made towards this result.

Stroan gages are generally only a means to an end, the
results sought being a record of strean discharge. Accordingly, each
gage should be rated as accurately as possible and daily discharge
neasuronenta (oftsner on flashy streams) made during the passage of a
great flood. It is often very difficult to make an accurate discharge
neasurenent during a flood, as the cableway or bridge from which the
noaouronento are node may be isolated by the high water, the river

filled with floating debris which is apt to damage or carry away the

89
current netor, and the weather cold and disagreeable. An accurate
knowledge of flood discharges, however, is of prime importance in any
flood control investigations, and the actual measurement of these dis-
charges must be accomplished regardless of the difficulties.

The U. 8. Geological Survey is to be particularly commended
for the mass of rainfall and streamflow information which it has collect-
ed and published on the floods of 1956 in northeastern United States
and the flood of 1957 on the Ohio and Hississippi Rivers. This practice
should be continued, even though a glance at the volumes of figures
elresdy collected nay suggest that such a great mass of information has
been accumulated that more is unnecessary.

Inprcvod methods of analysing rainfall and streamflow data
are continually being developed. It is well that these be given circa,
lotion among engineers to ascertain the advantages and weaknesses of
_ each. The theories of air mass analysis as developed by Norwegian
notoorologists appear to be a scene of determining the validity of
stern transpositions as well as indicating the maximum amount of rainfall
that might be expected over an area. The unit graph method of estimating
runoff from rainfall presents possibilities for determining crest dis-
charge as well as shape of the maximum flood hydrograph. hethods of
routing floods through reservoirs and down natural channels to obtain
the effect of changes in surface slepes and storagea is subject to a
great deal of improvement and simplification. The determination of
runoff coefficients from a knowledge of preceding weather conditions
and ground water elevations presents interesting possibilities; but these
are only a few of the many phases of the determination of necessary
spillway capacity in which the existing methods of analyzing the basic

data are being improved and new methods originated.

9O
SELECTED BIBLIOGRAPHY

Creager, Hillia- P. Possible and Probable Future Floods, Civil
Engineering, loveeber, 1959

Creager and Justin. Hydro-Electric Handbook, 1927
Fettis, C. R. The Probable lOO—Iear Flood, November, 1952

Cher-an, L. I. Streaaflow froa Rainfall by Unit Graph Method,
Enginuring News-Record, April 7, 1952

71st Congress, 2nd Session. House Docuaent No. 528, Tennessee
River and Tributaries

76th Congress, lst Session. House Document lo. 91, The Chattanooga
Flood control Problee.

Tennessee Valley Authority. The Iliwassee Fro: eat on the licensee
River , February, 19 59

m, mm ,e, 1'“ la Honk, 0.1...“ He utm‘., ”1.1 'e ”we
Certain Flood Froblees in the Tennessee Valley, lay 26, 1958

United States Geological survey. Floods in the United States, tater
supply Paper 771

United states Geological Survey. Rainfall and Runoff in the United
states, later amply Paper 772

lational Resources Coaaittee. Low Dans, 1988

lie-i Conservancy District. Store Rainfall of Eastern United States,
Technical Reports, Fart V, 1917

lie-i Conservancy District. Hydraulics of the liaai Flood Control
Project, Technical Reports, Fart '11, 1920

91
W

The author wishes to gratefully acknowledge the suggestions
and constructive criticises of Ir. J. H. Kimball, Principal Hydraulic
miner with the Tennessee Valley Authority; the courtesy of the
Tennessee Valley Authority in per-itting the use of unpublished
data free their files; and the assistance of Ira. lylie Bow-aster in

editing this thesis.

 

E TU'§_KY
IRGINIA
9's" "5931 if
”3;“ fi‘ 06
~ Iv
i .cfid
u. " at 9
’ ' ’ Rh .0

      
       
  

\\ _ .
raéii9$anfiia=
» new \.
~ WM’ 6573' Y3

’14

     

LEGEND

_ ----- — Slate lines
Tennessee River Basin
----------- Intermediate drainage areas

 

NOTE:

The IOOO-foot contours shown were traced
from the topographic quadrangle: of the area
prepared by the U. S.Geoloqics| Survey about
I900, on a scale of |=I25.000. New quadrangle
sheets are under preparation for this some
area but are not available at this time.

 

TOPOGRAPHIC MAP
UPPER TENNESSEE RIVER BASIN

 

 

 

 

 

 

 

 

 

 
 

 

 

 

 

 

 

 

‘ 7 I J -j

"T- I 1 r’ I

T FLOOD CONTROL INVESTIGATIONS
:Ml” “unseen vausv amoerrv

. I CITE. courses ”I: ”PM"
i files... - “I. h--- ... ................
g“, .“’ - pploIflxsgn

 

 

 

 

 

 

 

 

 

 

 

I

.mo.-._-----~“oo“m

E5

 

. Tracy City
Chattanooga
Dunlap

. McMinnvllle
Rock Island
Sparta
Erasmus
Decatur
Charleston
. Parksville

. Reaace

. Ramhurst
Copperhlll
i4. Blue Ridge
l6. Diamond
l6. Dahlonega
l7. Murphy

l6. Etowah

I9. Taillco Plains
20. Rockwood
2i. Crossvllie
2?. Harrlmen

3wps99eyw—

5:3:

NOTE:

The isohyetals shown represent the average
annual rainfall for the period I904 to I933. incl-
usive, and are based on the actual and computed
30-year values for the stations listed.

RAINFALL STATIONS

. Kingston 45. Newport

. Loudon 46. Waynesvllle

. Mc Shae 47. Breverd
Andrews 46. Caesars Head

. Clayton 49. Liberty

. Waihslla 50. Landrum

. Rock House 5i. Tryon

. Highlands 5?. Hendersonvilla
Culiowhae 53. Chimney Rock
Bryson 64. Montreat
Elkmont 55. Ashevllle

. Gatlinburg 56. Mt. Mitchell

. Maryville 57. Marshall

. Sevlerville 56. Hot Springs

. Knoxville 59. Birds Bridge
Clinton 60. Sraanevllla

. New River 6|. Rogersviila

. Mlddlesboro 62. Spears Ferry

. Tazeweil 61 Klnqaport

. Sprinqdeia 64 Mendota
Jefferson City 65. Bristol

. Dandridga 66. Bluff City

67. Johnson City
66 Elizabethton
69. Banners Elk
7Q Altspass

7I. Linvilla Falls
72. Gorge

73. Marlon,N.C.
74. Morganton
75 Lenoir

76. Caroleen

17. Jefferson
76. Mountain City
79. Parker

66 Dante

6|. Saltvllle

62. Marion. Va.
66. Wythevllla
64 Burke’s Garden
65. Bluaf‘leld

Scale 20 0 20

40 Niles

 

MEAN ANNUAL RAINFALL

UPPER TENN RIVER BASIN

 

FLOOD CONTROL INVESTIGATIONS

TfllNlOBtl VALLIY AUTHORITY
Juana com RANDOM eel-amen

 

 

 

 

. -é'fmm

 

 

 

 

 

 

TENN“ RIVER GAGES AT AND NEAR CHATTANOOGA. TENNESSEE

 

v

 

 

 

 

 

 

8a. Ill. (I. ”011““! crennsl and Cmtrol Accuracy
(3) m (6) (a) A" m
llateblishsdmw 0.8.”
W turned lo dllcm neeurs- 6.
s1 can ever to signal Service lions - n “m“
. 0.8 “mt ) 3. ll 1“ 1913 52.9%.,- in m of U. 8
Baden “:1?” P‘" d1
”40° 1676 w s‘am mfjo '0'” an. as 1902-1913 1891 ”up: ' m "m until
- . in cilr f
neersmdrssetin “maroon. polls
.‘L‘....“"".Za'“t.l’°l§“ .. ....,
vs c schar aeasur
31-“ tiabsr on cliff of left "m 9'- “ 1903-1913 1891- Mpln chem ot'u‘gfu
bank. Heatinr m.
81, m Vertical brass scale None 8-. as I“ 1913 lot reliable. Projectim bass

bolted to Pier.

of pier influ-lces accuracy.

 

Construction of tiles

hr Dc (:5 ailes

Bed exposed of loose
send rock,and gravel.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Autmtic recordim dmtreen began in hair constant. sun” ""38 CW". Ml rs-
sscflfiu 0" ’W I" U322, 21 m gage imtallsd. The 1905. in 1 lo rudro- a: t bank is high ‘1'“ "final” “mm“
“Hm ‘ ' of 1% is consi- aph coaperisans in- overflows M 1911 "NO-‘10 3880
“fill.” Mint 3" at "c023
erad standard. icets no backwater flood stages. tiles ‘ Mum”. "'1 "10°14
effect. 0coss No.1 Bar M is contr b‘ “"d '1‘" “W10"-
ol
capletsd in 1911. after Oct.&,1913.
Flow during low u 3 3.. an..." read
.9...” Welsh. U. 8. Heather mrseu stones regulated to (revel am Rock used tn” in,
h. a.” 6. 17 "all? 3'32. a: established in sale eitentrtu the about ale—mu sills“1 discrirgeuug‘isgroirggdpug?
1916 . operet on 0 power baloe . Pa '
a”: punt “ an“ a" see- r-naot. lienea by State Geologist in
he [T1 mlletin 36.
is in 2°ie
. . Results cmaidered ood alt
a?“ saw: an No.1 is 1684 1 an error in esti-stes'forw-
1.1916 um”"'.o‘. ”u.“ 21.aoo is vertical st’a’f rose Operation of power Gimme {tactically vidual «lawman: during 10'
a, Iiles above ant ' at (has in three sections plant cilanges slope gm Control is f1°"- ““1 curve “011 defined
«It-W . ”I... Cant! lo. 1 33¢ gm»;- on left or water surface. m0: planlth am new“. 11 tmam ass. ooo cfs.
nursferrsdto - '0'. ' tr tasenssubectto
as tan-at t “til. gag-fights. «tom to
Station not rated for disc
cat. 1.1913 “W6 nab-a ac coo stages regulem to drum and rock shoal Wflfiu‘mm 91”” In:
an. a incl 93: 33.8 3:: .11.- st's-a- "' 3- "“W WW- so. extant to open- about 000-le .11. 0mm min-t III-mom gap -
a 0-) tion of power plant below gags. Per-neat. “919'" “10'“! °"‘
at Hales Ber. mggmi‘23mgh-ith n
_ on pester
differences
used. iii No l is installed r10- duri l . l
. ng lo- ' 1hr
80 6,1”! :3 I”; ”tingle 1:: 21.eoo during 1 ,ussd with stages regulated to WI {statically thlz-mtmrmoem;
8 u m i. am 9‘ 0:0. Tilta‘rocmfl'. No.2 no. extent to opera- M'gfil 1‘ noting curve for th a condition
.1“. m t . . s c d n gags set to tion of power plant plant poorly defined. Records fair fa-
" sens att- es lo 1 at Hales DrL 90"“ ’ other periods.
. . ‘3'“) emu-01 for ”0.1.2 Slope -tM used except during
est 1 a. no.3 on laser .0 0‘91 lo.3 is a Bristol '1" duw“ 1°“ ‘9‘ t " “'1“ 3”“ mm m' 3'”.
' "" a raw... mi we a. . 1...... :mmt‘m "-0 5:3”? r"... “am-Wm .. ‘°' “°'
. . s a Bristol " 3 5
J.” .0. located on l a hour recorder. tion of poser plant gravel shoal ona- $31313”?ng
but _ permanent.
No.4. “Deco
magma m. m u... om... Mu recorder gags” gggglgym ‘0 ”:1? "3 mm "133 3"” in
M We c of ex opere- Sens as 92 - 923. sc c
’3.“ W tar mt 23,1924. tion of m:- plant 1 1 1 Power Cm since they were
3M4 “mun“-
lo.4 bolted tow (legs a .4 changed “0" “WW! 10"
”can” ”gnarl on 3%. Un- mrleyoas-hm neg-oer' "- :‘mlgeo to C°2§$°mtmgtu 1236.3
~.u'w m ”0'. mm in . comm 11'“! n 0”r‘- form [V "to” w 8. u Jmeto 309‘. m1.
ear pan-er Ml... stilling well. 3°“ °‘ 9°": 91"“ Dan, 22 miles below.
‘0 ° "‘ 7’ 21 coo (See Coll-l 7) 3.. CO1 a :2“ W... ' °‘
m 7 s of var at
a.“ (8-0 sects-sen ' I I (3“ “1‘" '7) will: maxi-ugh Sam 3 cwton;
eaaaee cm -
“ho med. IOJ Gags No. 1 staff cage to River at lseton.
u ‘ Maggi-ital 21 400 that Moatlgfi' “tut i "'1 pmuc‘n'
, e au out c pennant. Control is
m. 1.1“ stetiJassof no. :3 0:“ gagggrgr inwtia— a? “I“;ggllmm to col-pound. Channel M curves used; one a slope-
.S 1‘ ‘0“ above “In! (21 800 inch comgatezriron “a” extent by opere- 33:1 modified by velocity-.mcurve, “ the “Punt
0 street triage) no.2 at stilling well i - “°" °‘ 9"" 91"“ c "r cm“ W ' ' cm" mm c
cellar-loci: lo. 2 stalled June is, a n “t “‘1“ 3‘" 322‘me:3:2£¥°.: "11 “1mm ”com no"
at lee ear $30.2 is a vertical rabid; by water '-
h. cage. in s at power ham
M. 1.1“ Once No. l unchemed. Santeetlah coapletsd
to locations mm. ""' cage No.2 is a eeter- in 1928. Heterville so. as 1925-1926 mm S°°°° “”h'n“ “”-
cmged °
summer. completed in 1%0. sated 5" “7‘” ”fl“.-
osae :ust tele- titles 5‘
low regulated to
t‘ m6 “11:32" 21,” Hater-stage recorder. :12: fiflhyafifr“ Control is widows l I
dlllfllle- Mo
e I. ate non regulated to
u, ,, ,_. ml: 33?» ti: mam?"-
6 h.“ shag-but of 22.000 Hater-stage record-n power lent above guise-l is widows Reta-dd good.

30313" IN

 

 

s "1. m “-
acts strsal flo-

aftsr June 10.1%5.

 

 

 

 

 

 

 

~...//

 

anneal...“ at.

2. 00.5 o a and -n 54.3822

 

eeeoeaae 88.!
staircases seized. Jeeves apps! aliases
Pfizer—a: >345. Dunno-Ink

mZO_._.<0.me>Z. Jog—.200 000....

22“.: <oooz<t<xu .2 .3373... .25.”?

. c >0

52m mummmzzmr ag_33:_a.s...~_.s._w
mooodno 20.59590

umo >02 huo mum 03¢ 42. z 3.. >3. ¢a< «(a own. 2!.

ownun.nn~o_nn~n_an~n_nnmn.nn~n.nn~n.nn~n.nn~o_nn~2no~flnn~flno~

on on

en em

on on

N
1’

«a

D
V’

0'

G
on

on

3

3.

O
N

O
M

V
M

S
I..
v
9
3
v
I.
«J
H
v
I.
I.
v
N
O
O
9
v
.
.n
.S
.M
.9
9
v
OJ
3

O
n

s
N
C

3

 

 

 

 
 
 
 
 
 
 
  

 

  
  

 

 

 

 

 

ngno
0 TVA daily rain gages
' TVA recording rain gages
4" TVA radio rain gages
° USWB and private daily rain gages
° USWB and private recording rain gages—'“
RAINFALL STATIONS
”/
V A.
Laggno
0 Streamflow station K Y'
I Recording streamflow station
:1 Recording river stagestation
0 River stage station 3"“
T E N N.
3‘
f9
.. ' '2.’ TT
. ix
. DISTRIBUTION OF.
\ RWER GAGES HYDROLOGIC STATIONS
UPPER TENN. RIVER BASIN
FLOOD CONTROL INVESTIGATIONS
g nuance! nu.“ amour!"
A van. mum "PM
M'.‘t"“‘”“m‘

 

 

 

 

 

 

 

 

i
E
i

ii
52
ii
i

ii” 5 ii i in H H 2 :55 g

hoe-0nd

m

III“. M.

m.hl.
mm.n¢.

m. 111.
Vine-I00. ht.

It. ~cal-01.. n1.

III-m.“
hue-u

w. 1:1.
(10:?)

man.

aunt-10.111.

mo'm an

0. A.
5. In. a
w In. 20. 1903
“up an. 16. 1937
1.610 III-- 26. 1913
' an. 16. 1930
2.010 In. 1915
3.160 In. 26. 1913
' 105. 27. 1936
10.570 - hr. 1913
7.80 . tr. 26. 1915
' I». 27. 1956
0.180 hr. 1913
W In. 1913
11.000 hr. 27. 1913
' In. 11.. 1953
12.800 Ir. 2!. 1913
' Iv 15. 1953
15.310 Int- 198
13.100 In. 29. 1913
' 1.0 17. 1930
08,600 In. 30. 1913
" an. 22. 1957
19.30 In. 1913
33,100 hr. 193
In I! 12. 1935
" 3-. 15. 1957
1.80 105. 21. 1956
' h 3. 1996
as an. 1957
1.900 .7 30. 1!?!
' an. 16. 1957

awe-wavztmrr-rwm’é

t ”Menu-00 mrhfiam
‘I:mh_ammwhhmm.

3 Dad-math.

mmmmm-mmnmmwm

 

 

 

“p in at.

M r . Reference
20.0‘ 105.000 95-700 3.0. 100
15.00 10.000 23.00 up 025
1507 60.000 3509' ‘1’ 533
2.65 31.100 ‘ 18.60 In 025

-- 110.000 39.200 11.0. 100
2505 11W 3009' I? 025
17.85 63.100 16.9- ' _

- 135.000 29.60 3.0. 100
3209 Mm 1909' '1’ 335
350? 18.600 1009' "

- 110.000 20.80 3.0. 100
3501‘ 115.000 00.80 0.0. 100
310-9 ‘. 192.000 17.90 m 838
20.19 101.000 9.20. m 023
93.0 000.000 16.00 in 003
26.5 106.” 807‘ '

- 225.000 16.00 3.0. 100
20.3 00.000 110.60 In 023
05.25 mono 0.30 -
21.65 098.000 15.000 I: 025
35019 35.000 1000‘ "

.. 1.05.000 13.700 3.0. 100

. “0.000 13.300 3.0. 100
15099 0.600 2708‘ '1’ 323
M 10.700 25-30 "
22.16 22.000 11.20 up 025
21-69 19.800 1501:- '

500‘ 9.700 III-0‘ 'P 033
‘05 39.000 25.50 '1’ 825
80.79 17.300 1102 '

 

 

 

3.0.0.000 H can; r _ 3-..; ijxozx
AQRWEQN .N%.imw,. N. «QEQNNES -88 .8223 .8 8:00:63 2 9.33 2 .3 >2...)
83...... 83.118... anti-3 339.5% 92 E 5:00:30 3.550 Lo... moon 32.an 0+ 28

 

 

blulrxfio 05:21.0 Jozkzou 5.5:!
£501.23 >u....<> NUWMuZZu...

WZO....<0_me>Z_ JOmFZOU 000....

vooC xmmu .6 8228 2.0.8 m 8.865 8: £935 och
. 8.2m :3 2... 2.5.2 :38 9: 5082,.
05:6... $9: 583 355 note: 5298 c. 958% .6»

 

BESS 8+9 woo: xmwa 23:90 3.865 EEmmB mE»

. 30288 2 Emma taommi of B tmq

.mmzztoc of c. 85 2.6 of 52; Eat Emma .82”.
693522 9: 2 28m _E cm 000.com. L96 235 .3865

mu...<...m OMEZD zmuFm/‘u

 

 

 

muk<m LEOZDm 232_x<2 “to:
3:: mmgom z_ <u¢< “04.22....

$8.35 82532: 9:32;: 22,332: 2m: 8238:. 8552:: mzfi

¢m~ _ocem~ .moemm gowmm .mo¢m~ _ooem

oeono q-m N

on
ov

oo

oo.

952m 385:336 ncm mcfima bfo
Re .m~.m 5.. .6 ESE .363 .62.. .85 new 020
m8. .2-.. 552 .o 3.5.... . 3.9% 8:5 5232.32
mm: 4: 2% .0 E2. .fer 52

Re . 7m LmnEm>oz .6 E53 . ncoacu 52

92 .58 new: .0 ES. 3:2, 2.5
mmm_._mc2.-_m >32 .6 52m . Emma 52a 052333
32 . 3-3 :82. .o 6.5.... . £an L02: ncmtmneau
952m maoto> 67.3 .52.. ucmtmnEDU

mNa_.m~-~N 2822 B Ego? . Emma .52”. wwmmwccmh
mate? moot? . Emma 32.. mwmmmccwh

m4002>m

o§°¢EJ€+Dedx

m

o

o

mtz:
.V m

  

.....v§§.\.§ :féiu

IUU!.OIU

 

 

dozélauTtxoToelTQoH so.“

 

 

 

L...
.

_._.

 

N _

OOGDVMN

on
ov

on

oo
oo_

com

com
ooV

com

com
ooo _

TIM 3UVODS 83d 1333 GNODBS NI 398VHDSIG

ooo~

ooon
08 Q

 

 

 

mm<m> 2_ 55:05...
ooo. coo coo 03 com com oo oo 3 on om

rivi¢0ua 01.83.30 02.03251.
5.7.3....qu raj.) 51.327:

an)... unguZZuh

 

.ZZu... .(OOOZ<PP(IU
mwfiuZuDOmmu 000...“.

 

 

oov

 

 

 

con

 

 

ca. '1. mm - man“:

HO 011110.113. Inuuuncm

‘

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

lPro'ecledon ’
probe ility cmc\ , ’ fl.
\ xiv i"
sum ’
\N v/ /
J '/
JV
ILA"
49V
0 4"
E 4 /
3""‘m
_¢_: A.4fHaxirng’rg‘gecorded
5
.2
3
O
gmum .
.5 ~ 00m 1080010
2”.”
/%
100.000
3 3 § § g g g.
-- 9.

Probable Frequency in Years. with which given Flood will be Equalled or Exceeded.
Flu. 27.—Frequency Curve, Tennessee River, at. Chattanooga, Tenn.

From Hydro-Electric Handbodk

by
Creager and Justin

1 927

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.5
Graph showing relation between the and ratio of
§ miner: flood to aver-are flood by mller method
a 1.0
‘ l
a 1'5 I I or
= 191' tv .| beSt (it.
3 2 o Jpo can” or b I
0 ° ‘ 93? fit.
2 H'
.2 2.5
n
O
3
018 3.0
u
“3 3.5
“1 lllllllllll
Reqmoy curve of annual floods.
\ Tenneeeee River at Chattanooga. ream. 1876-1931
1.8 x,‘
\\
s§ .
1.5
8 \
o e
c '.,\
Q
3L2
e
E
= 0.9
o
0
o‘
I:
«I 0.6 K
Fostc type I -—-— \fl
meter type 1!! K P ‘ ‘
mm ‘—-——- N1
9': ‘ \
N \
0 o. in i_1
§§§§"““28833328888m$;$g
0 0e 0| 3’:

Percent of time in yeere

Frequency curve of annual floods
Tennessee River at Chattanooga. Tenn” 1875-1931

hater type I —~—- -
Poster type III
linen —————-

Reno to avenge and flood

       

RRSSESSSSSSON”

0.5
0.2
0.1
0.01
0.001
0.0001

Pement of time in you-e

From “Floods in the United States“

Tater Sunnly Paper 771

10

M a.

e
O“

2.3

'89:.

28's

6
p

Type II

J

'.ll.' FHA l l

Latlo tn

Function of R 'lth reepect to tine percentage (IR)

doodrlc. type 2: frequency curves for annual flood.
on the Tennessee Diver at C attanOOfa, Ienn., 1875-1931

tempered curve

t approximate curve

line of beet
0.1
in
33335 ~ Na” 2 8838828838

Percentage of tine 1n yeare

noodrlck typee II and V frequency curves for annual floods
on the Tennessee River at Chattanooga, Tenn., 1875-1931

tralsht lines of heat f1

78:.

'8

'2

.001
>1 a?” ~ “a" 2 8838888838
0 o 0 <0 0
3% Percentage of tlae 1n yeere

from “Floods in the United States“

Water Supply. Paper 771

 

99.9
99.99

99.9
99.99

 

R e Ratio to average annual flood

 

hue to nu fleet

 

 

 

 

latte to lean need

 

can of annual floodo, temeeeee River at mttmop. 2a..
1376-1931. at. by Blue nethoa

baa annual flood = 2m,560 mend-feet 0"“ It“. 0 0 0
31‘. curve —

8 ..

”teen-lame

method; renneeeee River at Chat

1 - taller fox-mu

B - Inner I-thod, beat fit

c - router nethoa, type 1

D - Poater nethed. type In

I - aaaen nthod -------
D - Goetrioh Intact -- —'-—
G - Slade lethod -—--—

50
100
500

§.

0.000

The 111wa

From “Flooda in the United States"

Water Supply Paper 771

 

.

 

 

 

 

 

 

 

 

 

 

 

 

 

 
    

 

 

 

 

 

 

 

 

      

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

       

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. 4 _ . . . . ..
a.“ anx a. 3: N U 00“. Sex 23¢ \KQQU \xxfificbaus {mafia . (9.000%
e I e I'll . ease! 3...}:
I e ll. . 4 3a.»: eaeeIe-oeeu
O - l , C 0%.. we: sate-e:
e r l | e “V0! V iii-.09.
Qua lifit bad ~§r€ f 'a 30'. .‘ veto-a.
OZUQ 4 ._ \. NER \0 NBVNKMUQMO.
90 a 90 a. Go 6h. 2 9‘ h. QN HQ ‘5 “G MD \Q .8 \Q
a , .1 < 44 l 11 4 A l 4 . A 4 l ,4
S“ .‘Di‘. ‘ L'-. 'OIUk o v on .n o a e. . .. a ... . . o 5 . o o u M. o . e o 4 . V . u a . o . . e
wu§0.203¥.o8dug¢e .. ... .... .. . .... .. . . .. . _ 3 . . . .. . . .
. . .... .. . ... . . .. _ . . . . . . . . . . . . .
biggttdb .. .... .. .... ... . . . ... . . .. . ... . . . .
Q ’6 irizlrtfel l o f 0| ' lei] e e 1 o e 0 9 a is 0 e 9 0 l [Ill 0 . 6. +1 lit I u 0 it Q 0’ (1| tillJ‘\°
”In! & a ”W . .... .u . .. . . 4.. 4 4.. n .. .
. . .o . .o .. .... . ... a a . . . . - a . . . . e
. . Wufi. . . a o. . .e . .... . a. .m o a o . . _ . . . o . . . . .
. . . . . . . .. . . . .
1W ”FYI‘1Lw0lFl‘Q “-0 tig'fi b V k n .* h e e n e D e D it e I a v < * i a .o M e 0 lo V C O.|.O ..l ' -l-l|lI JAN“
. ... .4: .. ...... .... . . 4 . _ . ._ . . . _ ,..
a . .. . . o... o .. .. .. . . .4 .. 4 . . . . 4 . . . . . . o
. . . . . ... o . . . . v. . .< 4. . . . . . . i . . . . i . a o . e .
Y... o .9 o . ve . u .o . . .. .4 4. . . . . _ . 4 H . _ . . . . . a _ . . .
It- 4 F 11!. O.- Ocevltlele Qlilo e e e e a ...: o e o e e e o e .. l... 0 r1 0 9 o . I ll 0 I. 9 no If e o «i a. 0 e I I. .0 i- 0 a 1 IO f..a
T e e O. a 0“ Ha e e v e o. .... fl. .. o o v o . o . . o . . a a . . . . . o . + . . e . e . 0 v
v al tee e e ev... o a e. .a c . . .r “9H. . . u . . a v . . H . H . o . o . . ¢ . H u o .
-01 Qvlll. ‘1. ...". ...- ... . ... o . o . . . . . a . . . . . .
..HH. ..3..1.... . ....m .33...“ 3.. , . 3 . . m .
I 10' ll- 0|.» l0 .0 l o o i... e l a i i i :0 I0 I i o o v §.ia l .. l in. .l 9 - lo I '0'. [bit I ll
v e e to a . . c n. o e .1 ....1T.’ . W Hfi+0el . . v W (0. . u . w . M . M 1 e . Ifl . N . e. . H . a
Y.e.l ee . 0 .wt . v a . . . . o u o . . . . . . . , . . e . . . . i . .
We .4.” . . m . .. . . . :. . m . . . fl . . m . . . 3 m
YI‘IOJV" YIFLILIPhe 71.0.1.0 0“ lo. et...91r.'llta k ll n e on n . a w .. o. h .v u g l i n i a . n u . . . k i . u n t o i 18
e e . V F .... n a o . . o . . . . 9 . H 4 a _ . 4 + . 4 . o e .
IlliI Ill; O’IAV-....-‘IiOlela“..t.¢000-1v'.ei6t O ‘0 e V O I o o a u e 6 e o a o v . l o v . O o + e o o o a i e e o o $ 0 O I o 14 'l‘ 0 I? i 3
e e O . . u. o. . o .. . .. . . o k a . . . . . . . . . . a . . . 4 . .
. Il.v a-’zlvlve.i¢lll.¢.v.o i.v. . i . o a . . . . . . . . . H - . l . . . . v - ¢ 0 . o l. .|+ wit all i- k 9 l o a i .NQ
. i - ee ... .. .. .... . . u . . . o e . 4 . o . . . . . . o H v a l o v o a a i
1 .Y'COPAYll’tlm e 0.1. e 670 I H. e I it l o e v n i t f o u o b a A 0. b e. ... 6. i111; *lll-|i§lll0| O'Trllllaa
a 9. I i e o .. . . u o . . . . . o o e
I:.v+llf{9+llvv0lzvv¥ Yliao'e.o i h h h e 9 'le 0 . . . e. . o i u .9 o i 0 fi 9 N. . m . .4 9|- e ‘ Olive. 6 H o l -Iii ‘8
Pnliwbl .oil-...iw»i.1ll-l....oh n“ 3 n . --w a . . . . a . a . 4 - - - “ - .il- (I. i. -41 l? -L. its. Till; iloni ll .9.
. . . .. . . . . . . . o . 4 . . . k ,. a
. . . . .... . . . . . . . . . , . . .
o .. . .H. o. . . e e . . . . o a . o . t
H. . . .. . 4 .. . . o . . . . . . . .
.1..le [filer’ rTi' e: r O 0 V i” l b l v + 0 VI ’. 0 9. . § it ll- ’1 911! .e l ’ lily 0 a“\.
om . . . e a o . u o . . . . ‘ o . a o . u . c . a o r
. . . . ... . . . . . 4 . M . . . . . o n a . A A
.. . . . .. . ... . . . . . . o . \‘I . . . . . a . a
...?rrarl- .-.- .-.: i-».-“h..h- .. . - . ... .15.“...3. . p- n . ...)- .. ll 7|- -«Il--o-+ r.~ JV
.. . . . c . .. . . ‘. . . I ~i§Hul~ \N., . . . . . . . . 0”
.e . o . .+ e , . 4 . e a . §.‘ .0 ‘0 w . . . a . u . . 5”
.. .o 9 . . . . . ‘il on u C . . . . . . o . o an
.o . a . .. . a O . . . . . o . a ,
0"16oln. e v . u .i a‘. i o l V a e t i.. l l v ls! Flilillll‘lllllfil r!
a . o . . . n . . . . . e U
o . H .. . a _ a . . . U
. . . . . . . . . . . .
.uwp--»-..h». . . .p 3 - n - .. . . . 4.- .,-l.|..tllv. m.
1 y a u c J n I . u e I I
. . . . . . . . o u
w m 3 . H . u u . 3 , . n . . P
O.» O i i a. e. o e l e I O 4 . v I O a a O till i“-
. . ..4 . . . . . . .. . a 0
r O .Yeeeth‘Oee.. e n e v i o .v o o i i v . it A ll. 1‘
. a . . . . a.
o l a. o o 7| 9, . . o v u . i . o ‘lllh J
. . . . . . /
a o a i c o e v + o . o . e i e 0| 0 o I Oll1~. x J
. . . . o . . . . .
3 3 . .. . 3. . .74-”- ....- -- .3- line #5
i e e e 1 i ‘l l ' -1 ll? 0 IIPI" i 4 I l.’ | i I- l T‘IIOIIIIIA Q.‘
u n . o o a e O I. O f O O a I C t O O 9 A
. 9 4 . o . . o v a . e < . M u . . ~ 6 a u e .
. . . . . _ . . _ . . . a . . o o . . e o e v e e
. ... N . . . . . e . . _ a i . . o .- ¢ 0 e e e e e
e- i.» 7'0» c. ll iii 0 e u i e o O O b 0 ¢ . a it I'll ‘ *Ieli‘ ’ I'I } 1 1
. . . . . . % o . . . .v u . . o o l W. n e e e e a e - “fl
. . » . . 4 o < . M . . e . o . u a . e e a e e a
. . . a . _ . . ‘ . . . . o . g . t . . u o o o
--.-..fwi---“ ...»...1-.-L. . Til-..“ . a . . 4 . «lllb-i ”he 4 .r -..; . i. .. H . ....
. . . 4 _ 4 . o , . . . . o . . . e . e . e lo ll.
4 . a H o .. o . . o . . o I ie 0 i H e 4 Q o e rIlY Alto 'l o e 1
. M . . o .. . . 4 _ . 4 . . u . o 4 o o n O o O we a e. .
. . . . . .. . . o a . . o . . . . e a . . . e e e
vilelt I ’6'. e . . u a. .. l'oie. ..00 a r e 4 . o u . o o a i . i L'Ii‘l- I'lif I ll > No
. . . . . . . . . . . . . . . . . . . . . . . . o .
.. o . . .. . . s 4 . . o u o a e e e 0 e e e O. e
«. . . .. .. 4 . o e g i o . e a e e e 0 e 6 e
Y‘rhvo of. “e0lo eh ll; 06. 9. O-Nhg e e e e e t a a o a I 4 . M A i 'l. . h Plirl lr‘llArle’l all! . v e e .h e e H .v
c. . . . . . . . . 4 . o . . e e . o
.. . . . . . e u . e n a . a v e e e o l
a: k . 4” . . . - H . 3 .. 3 . . 3 u n h . . n- .
7.00970 l e e it ii .0 O '0‘... I000¢ 0 e O V e e e C e u l e e a e 'AY'OilFlIIOI‘I- . I’ll: Olli4|1 M Mr 1 [Ml ~
.. . 4 o. .o . . . . 4 . u o . . c u . o . a a a 9 e 9 e
{9911... .eOOOIAv‘O... vttlllvloo’r-l “’9 Av e. e .e l O .0 e o e e 9 a . e c a e I o 0 . lIII'IIOI‘leYav" '57 .r 4} 0 «r I“
. . 4.. 4 .. h. . h 4. . . . I . o . . . . 4 p e e o A 9 e e .v .
.TIO: 090210;.Y90991 9191 Io e . e e . . . o . . o o a c o 9 l i l 0 cl illILr'OIol l t rr 1 B
.H.. .. . . .. .... . .4 .. . . . , . . vi. . . -. Tl o . . . 4 *Ji...+.+ . l I
%l '0. .‘I 14 } “11> +0.. 0’ ‘l 0 e 0 e 9 e 0 Oil. 0 IO! "Il .0 0"- l9 t [L1 4 11 r} t 9 in "
OH fl . . .e .. . .. e n e o b . . . . o e a e e a H 1H w 0'.» +0 I vl rl. e.
l 0. 'l' “w ' ‘ 0 e 9 e e e Ie r V 0 I- V .6 i -l"- ‘Ol 1 1 1 4 1 '
mm): a. a y»: ». ...» .~.. ..n» m.»» . . n .. .l n . l. . . . L. 9:? . .I » Ti W- ..4- .IHFL. 1..
O F 9.
>
.
. .
o . C
O o o
.
t, I * it'll, '

 

 

 

Ev. NU§§K§Q g §qk
Q‘NLiSn. ENEVQ gut.

 

 

._.+ .—
‘T” f‘ ‘f" '
—-—+——- +-e f--o
9——H~-o~ev—e——Tro—e

*-
..+_

 

lisgag

 

 

Duke's awn. S QUKV «$.00..ka

 

o
O

o

6
e

'

3

efee.e

.ep
vie

Hz-.. ...! .

.
....
e-‘0

e e

o
4i
1
a

 

 

Ozel+ll+l..

\V\

“x

 

 

P...
on
S
90.
m
c.
9
0
9.. M
8 a
4 d
a
. J
4 c.
. ...
H w.
W. m.
. .9
.4.
n ,.
Q8
8..
one
one.

 

king‘s at)

 

 

 

 

 

N0 TE .'

Zones shown hereon ind/bare fime of
run-off frave/ ac compufea’ from fhe
sform records of Dec l932.

S YMBOLS .'

— " ‘ — Sfafe boundary
_ Basin boundary

— — — Tribufary baa/n boundary
—— 24 hour zone boundary
------ 6 hour zone boundary

Sale l0 0 IO 20 Milee
L . I I

 

 

RAT IONAL M ETHOD
TIME ZONES
UPPER TENN RIVER BASIN

 

TENNIOOEE VAtLEV AUTHORITY

WITCH CONTROL PLANNING DIPImINT

 

 

 
   

322 333“

 

 

 

 

 

 

 

 

 

 

‘or o

l

 

  
 

  

AY
l

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

  
 
 

 

 

 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

—J i
:3 l t
i 70 . - 4
O {MEAN OF 60 YEARS
2 L---— I. _ l
x...” r
Pm 60 T 1
(—d -
‘3: ,
“1‘9
\—
( ~ ~ 4
G.
a: 4O. : r——-—-—-J T — \ ”‘4
5 »~:~=;;l;’v t 4 - lAVERAGE TEMPERATUREF ‘ l ,\
*- 3or—— l ‘ 1 L l l l N
6.0 T . i i I 7 . 80
~ 4 «l -$ 4 I i - + l l - +
5.0 lmixnl 1 i/J‘KRA‘WM - -70
: l I . P l E
LLJ
f.-- __- . -.-. g
0‘5 g:
3:
Q 2
E '—
La.
‘5
2
D
m

 

 

 

 

 

 

 

 

 

 
 

 

 

 

 

 

L
~
---

" R'A‘i'Nfi’ALL AND RUNOFF

 

 

 

 

 

20

 

 

 

JAN

 

 

FEB MAR APR MAY JUN

JUL AUG SEP OCT NOV DEC

 

 

METEOROLOGY
TENNESSEE RIVER BASIN
ABOVE CHATTANOOGA

 

FLOOD CONTROL lNVESTlGATIONS

TENNESSEE VALLEY AUTHORIYY
I"!!! CONTRO‘ PtANHING DU'MIYIHNV ,

V"

 

 

 

“ ...QI‘IVIID I ' U. 9'0 .‘V‘W'IC4 ""
wan—W» -
KNOXVILLE ll-l8‘33 Ev LPPi 0 I lA|32R

l

 

 

I;llffi1mm--_._

 

 

 

 

 

 

 

l
l Sale 0 0 I) m um

 

 

 

N0. / CUMBERLAND FLA TEAU
' JUNE :9. 192:

 

 

 

 

 

 

 

 

 

 

 

 

 

     
  

 

 

Ham
6' 33% Alto
/\
w ‘55 /
N0. .3 WINCHESTER SPRINGS . N0. 4 BURNS VILI. E, MISSISSIPPI
JULY 6, am AUGUST 29-30. I930
W/r 0.... WM...
' ‘ “5° m
0". J ‘29:"- Sf Paul
‘W, ' A ‘ cum
I" v»
(”man \ /
mill.

V C°"¢9¢

I

 

 

 

 

 

No.6 CEDAR CREEK
NAVISJO.”

 

NOTE:
77)::farmsshom moo/kwdfobcrypia/offm
summer Mandel-sham and cloudouaf: Mo! m

--' \\ \\\‘ ~'--- 4‘ A I"
\ ‘;.~' \“ \\" ‘ '
\\\-\ ‘ ii 6» ,/ ,i’ like/I’Lfo oczgr eggslfnwf any sod/m of Mo

5cm 4 O 4 8 Mile:
CE‘ZEZE

Except as noted

 

TYPICAL STORMS
CLOUDBURST on
._ THUNDERSTORM TYPE
\ - TENNESSEE VALLEY-REGION

A L A. .
\
L004 770” ”AP F [“0043I ”(3‘0 NTROvl‘ligvAE STlGrA'I ION 5

ram com mm WINCH?
new Am

HIM. ' -
r mom], a-a-uhfipplnl 32262lno '

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_ ru ”-00.04 7 ,__

 

   
     
  
 

K Y.

/ /
f "" 7/
/
/
/ ow;
$
’\

‘ jewporf
”39005 .2
Ce

0’

/

I

”33

l 2
m
1

Asheville 0’

  
 
 

 

 

 

 

 

NOTE:

The isohyefals shown represent fhe thbes’r
rainfall center resulting from The WesT 1nd38n
Hurricane of July I446, I9l6. They have been , V [A W ‘11 .1
plotted from dafa published by +he uswa. «car '5 i P. :9 , 4E " ”‘5
The high eievaflons nOrfheaeT Cf Asheville
re5ulfed in The inTense rainfail bemg Um—
cenfra+ed in This area.

 

Tppoqraphy of The area was carequ‘ "on- D'STRIBUT'ON OF RAINFALL
ssdercd :n defermming The shape 0 These
lineS. WESTERN NORTH CAROLINA

.7 _ STORM OF JULY l4-l6,|9|6

 

FLOOD CONTROL INVESTIGATI ONS

TENNESSEE VALLEY AUTHORTTY
*‘an (IIZNI‘IOl FLINNIN'. 0| 5'.” UI'NT

xvainuu :( ‘ Ann .
tau-“l I‘nupuuo ' ’ I" ’1

...“, )g’d cod/JD 7“ M4 414,;
a.“ PM KNOXWLLE

FU 33-9: 63 i"

ll! . ‘ ‘7
'0’.) 2n: sun: ("Injury INS?
J L

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

“I v M
ILL no 0“” w m
I” N u
m
min nc
m I“ sc
5
GA
LA
m
SEP 4-5. |9I5 JUL S-IO. |9|6 JUL I4-l7. |9I5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

OCT I6-l9. |923 AUG IO-Il, l928 AUG I4-l6. I928

 

 

NOTE:

The hurricane paThs shown represenT The movemenTs
of The low pressure areas which accompanied The
disTurbances .

The maximum rainfall areas shown ropresenT The
IOO-mile wide zone of maximum rainfall which
resulTed from The passage of These sTorms .

Basic daTa for deTermining These paThs and amas
were Taken from published records of The U S
WeaTher Bureau.

 

 

 

 

' Scale 200 0 400 800 Miles

 

 

WEST INDIAN HURRICANE

 

LEGEND
“ LOW PRESSURE MOVEMENTS
we ”mm '°‘"*°"a'°'° AND THE CORRESPONDING
-——* Hurricane pa+h RAINFALL AREAS
@ Rainfall-inches FLOOD CONTROL INVESTIGATIONS

TENNESSEE VALLEY AUTHORITY
wan. OONTROL PLANNING DEPARTIENY

 

 

 

 

 

 

comm-to REOOIIINOED “ ”mm
MAW (I ’x‘ U. BuilA
KNOXVILLE io-A-aa] w IPP] o I IA68R:

 

 

 

 

 

N D
S D
w
NEE
RAN
' OKLA
R‘
‘Av‘fvnv/ -

TEXAS

NOTE:

The arrows Shown indicaTe The locaTion of
The cenTer and The lengTh of The rainfall

zones of The ouTsTanding sTorms which

have occurred. or which if is believed may
reoccur, in The viciniTy of The Tennessee

River basin. DaTa for This sTudy were Taken
from The publicaTions of The Miami Conser-
vancy DiSTricT and OF The U.S.WeaTher Bureau.

  
   

Ia cor .cf

 

 

 

 

 

Scale 50 O ICC 200 Miles

 

OUTSTANDING STORMS
OCCURRENCE AND PATHS

OF GREAT RAINFALL

 

 

 

 

7.,- mm. ' FLOOD CONTROL INVESTIGATIONS
' ‘°‘ °'“‘ W" "‘"- TENNIS“: VALLEY AUTHORITY
...” "I". CONTROL 'LINNIM OCPIITn'NT
‘ “'Nn“ auammo "consume“ APP-oven
"m” .‘I‘..‘..‘.'.;'*“ Tag-4a: -, 14;.-- .:Q%
a...“ KNOXVILLE II -Iz-37[ STOIJ OJ OAZO RI

 

 

 

L;

—r

 

“mun—— ‘7 ...—7 .777 77

 

 

 

   
    
 

M0

    
 

9%“

”A

fiat‘imw 9

    
 

 

 

 

 

 

 

APR I-S. I920

 

 

    
    
 

  

Wm

 

 

 

 

 

 

 

MAR ZI-24, I929 DEC 24-3I. l932

 

NOTE 1

WeaTher Bureau .

 

 

 

 

JAN 20-25, I937

LEGEND
§3§X ' M ' ' f u
S \§ § ammum rain a area
7 PaTh of low pressure area

O Rainfall - inches

 

7‘“ ...............................
m... m .....

 

 

The paThs of low pressure areas represenT Those
ThaT accompanied The cyclonic disTurbances shown.
The maximum rainfall areas shown represenT The
IOO-mile wide zone of maximum rainfall which
resulTed from The passage of These sTorms .
Basic daTa for deTermining These paThs and areas
were Taken from published records of The U S

 

 

 

 

 

DEC 26-29. I926

 

 

 

 

 

MAR I6-I9. I936

Scale 200 0 400 800 Miles
_=—===:=l

 

CYCLONIC STORMS
LOW PRESSURE MOVEMENTS
AND THE CORRESPONDING

RAINFALL AREAS

 

 

FLOOD CONTROL INVESTIGATIONS
Tennessee VALLEY AUTHORITY
WATER come EARNING OEPAmENT

 

 

 

mom: lO-4-3I w PP OT IAGORO

 

 

I'll 1-nn n9

 

    

 

 

 

 

_. oan_(_ 70 (".7 373-9-.. uijozz
. IVQQNI 3. ._ essay . .
H 33:34 ouozmusbwuhllr 35.103 03. oiv .50 005?. ><olm ooa. .NI: >02 ooEum ><oin

 

 

FZU’Ffl‘flUO UZ.ZZ(I.1 JOEPZOU KHP(3
>P.¢OIPD( >UI_I_(> UUWWUZZUP

mZO_._.<0_.rmu>z_ Jomhzou OOOIE .seaeieeazaséaalsi

H tom n32» 3:5 :35 3 =93.
ER? .5: if. to: 522 E8... :8
“an!

.. .
a. xi

:28 35528 25: ..taa 52!
H .5. “8.5. 3.5 523 s :95.

5.2m; SEC :9... to! :83; :39... 28
O H E!
, 12/ L
\l/ll 7N
p. I l f 4. pl.
__ . _
, . - A
\ /. ..

 

0.0. OZ< mom. Nam. .mmm. don.

 

I) IE

Ij<LZ.<m EKOFM w>_mmu0xu A

F w

7|) ..

   
   

  

    

 

 

 

 

 

I a. Enemfim

com o oo~

 

3:: 8a; 28a. :3.“

.3... no.3, 8885» ...: Les

hau- nl. 5....) Etc: 2. a 2163..
.35 .5... e .28... :2 25: as:
HEb:

So. .07». «<1 ooEua ><oI: mac. .ONI: Qua ooEum ><oIv

aces Eggs .5... .68.; .353

N Ea ..aEm 33 5:5 S .523.
E2». .8: a... a! 822 $55 28
”ME!

323 :52: n: a 8.8.. 33:8

 

 

 

 

52.2; «<2 aoEum >3;-

gao ifl: 5. use 3.333

 

 

        

moo. ._ 22?; ><1 no.5; ><oI~

Reno $3523 .5... .3 .328»
n t... “3.5 3.5 625 S 5......
5.2? ea. 51 :2. Se... 5.... :8
up.52

 

 

«an all.

 

 

 

 

 

o¢0N_<_ 70 Ta; 2,—3-2-.. uijozx
(«IQ «CRT is ,WMEVIRVIQ masses. . .
33:: 30538: 8:33 32 2-3 «<2 SEE EA EeaA >02 8.5: 3:. :2 n 52-3 SE 8.5: :3;

 

 

 

P2325230 02.224...“ JoxszU 5:;
>P.¢OIPD( >UI_I._(> UMflWUZZUP
mZO_.—.<O_._.mw>2_ JOKFZOU 0001....

 

 

 

mmm. oz< Km. .22 .06. .mS.
I_._<.._z_<m 3.8m mammuoxu

 

 

 

 

II ill . IHSIHHHHAJEII

. -. A T
3:: 000 Sq com o oom 33m

.3... as; .3855. at 36

38.. he... 5....) «Eat It 2... 9.95..
.852 9:3 I 38.8 8!. 252a 82:.
H who:

22 .oTE .5.) 00.5; ><oim

guano EESSU LES: ..taoa .825.»
u re ...ezem 3:5 525. a :55
gm: EB; :12 to! not»: 5.9? 25
”ueoz

.an .. 35252 5399mm
E :68 «:32 Co 299: 2688» 2: ca

22 .nmlmm «<2 oozin— ><oln

aces €2.83 .25.. .925 .82.?»

u E... ...asa Sec: 525 s .55.

m E83: E9: c3: 2:: care; 56;... 2.5
O H~62

 

I. 3855 :5 .0 ream case I. S
/._ :83... 2.: 8.5 :83 82; 882 56% £8 ... .252... ES. :93 to) .882 98.1 28

. .
c, We 7” .. J3
\ .x N
HILIT. .T .u . ml 7 .V. ..
7 J. E ,

...Omfl c.9533

 

u who:

\

\

473/ El... - KICK
U I“

 

m3. .mTo. 25. 09mg ><oim

.tcna $8.95 tax .283. .328»

S .5. been 1...: .55 s .325

Eben: 5°C in. I! g CIR-n £00
N”to:

 

 

 

 

 

oth_(_ _o T: 373-019; misxozx

 

..nn‘x‘t ‘x A. M $.\ “0“..me N. I km»

33...}? 3023.382. 3:59;

 

l;

 

 

 

#233563 ozizjn Jozrzou cut;
>P_¢OIPD( >MJJ(> UHWMUZZUF
”ZO.F(U—FWU>Z_ JOfiFZOU 0004:.—

 

52 oz< one
jEzZm. 3.05 35888

 

 

 

 

Tdnuflfi W I: wanna».

' . U,
8.... 2: 8a 8w 0 2.x :3“

 

...-:8 2.1.. .0: .... ... €82 .2an
En: 1E8 to: 822 E65 25

.3..- ao:-> 882E.» ...: .26

53.. E coil Etc? .5 3 33:5.
1915 I! 3 3.8.3 to: 2:8: 32¢
..Htoz

c3. .o~-o_ «<1 ooEum ><oun

 

 

5mm. .nm-~_ ZS. no.5; ><o-v_

\

- 4 L \aK
./ .. :\W§ .‘v )
. ... a ..\.v. 3&3 _.
. w\ cw. - .

(l’I

   
  

52 .nNuc 25., 00.5; ><o-o~

I‘M} . . 9V
, _ «5V...
....u...\\.\.

\ “N93?

wula\ I

 

 

'4, I869
I'JUN I, I889
' 4, |927

MAR 22 ' 23, I929

Z-DAY STORNS

3 4 5 6 T 8 9 | 2 3 4 S 5 7 8 9 I
HUNDREDS ' TIDUSANDS TEN THOUSANDS

IN INCHES

3 AND 4-DAY STORMS

3 456709I 2 3 456789I 2 3 456
HUNDREDS Tm TEN THOUSANDS

I
....
O.
U
D
{‘9‘
<
I:
U
>
(
53
t‘.‘
<
U
a:
0

5 TO 7-DAY STORMS 1 DAY muoo MAR I6-22.l936

3 456789I 2 3 456789I 2 3 4567
HUNDREDS 1mm: TEN THOUSANDS

STORM AREA IN SQUARE MILES

NOTE=

The Timo-Ares-Depth curves shown represent

sIorms which hove occurred over toss in sos'ern T'fiE - AREA . DEPTH CURVES
sod northeastern UnIIed States. end M hose run—

ffchsr ter'flc I IIsrvoI’hosedIblhpev-T -

:ssoo RTter L;.s"1"hese curves Ihsl'eforo giveem EASTERN UN'TED STATES
sh Indication o! Ihc rsinfoII which my my be

expected to occur over this ores. Storms occurr-

ing slang m. South Am“; end The (3qu of you“ FLOOD CONTROL INVESTIGATIONS
cows hm hm cchod-d- mm'mm

DsIs for these curves hove been Isl-en from published

dsts ofthe HIsmI Conservsncy District the US

Westhor Buresu and fho TVA.

 

o-al-ss w n o IA33nz

_\.\\ \

\\\\\\\“.

“i {cu

‘ “c1
‘ 1!:

 

-_- --.‘-_-.u~ - -...... m":— APP-‘2

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

500 500
459.000cfs
3“) 4| 000cfs 39!,000cfs 400
U __
8 341000075
5
t5”
0:200 200
<
I
U
0')
5m I00
0 0
I2
U)
U
E 0
Lu ROM)“ RUNOFF
2 RAINFALL
’2
..I 4 RUNOI’T
D
2
D
U
05 I0 IS 20 25 30 20 25 2 7 I2 I7 22 27 26 III 5 IO I5 20 25 28 5 IO I5 20 0
MAR FEB MAR MAR APR FEB MAR
I867 |875 I886 I9I7
$300 300
U 298.000cf5
247.000 cfs 250.000 cfs
§ 2387.000 crfs
-200 200
E
m
3100 I00
<
:1:
U
‘_’3
Q 0 0
I2
(I)
Lu
I
U
E a
L; '* RAINFALL RAINFALL
E
5: 4 HRUNOFF RUIIOFF RAINFALL
2
D
U 0 0
2530 4 9 I4 I9 2025 304 9 l4 202530 4 9 I4|52025 30 4 9 I4I924
MAR APR DEC JAN MAR APR MAR APR
|920 l926-27 I929 ‘ I936
RAINFALL AND
RUNOFF GRAPHS
PRINCIPAL FLOODS AT
CHATTANOOGA. TENN
FLOOD CONTROL INVESTIGATIONS
TENNIIIII VALLIY AUTHORITY
UATIN CONTNOL PLANNING DIPANTNINT
9mm W9 IUINImO 0 COIIINDIO AWN.
"...... -------------- mum KW v
°m~mmww ----- momma” II-Io-aalwrpplol IAI24no

 

 

 

 

 

 

1...).-m

.31

 

...Ian ’l—. 75..an

11

'

...».iEE 3 7

 

 

 

€393? 7'. "
”‘7" .fitl‘. H

I
......

     

NOTE: . ..
solid isoh etal lines indicate the posmo-n
nghe Ohio Raver Valley storm of March 22 27,
l9l3,ss transposed over the Upper Tennedsse;
River Basin In such a locatlon as to pro uc
a maximum crest at Chattanooga, Tenn. th t‘ l
The dashed isohyetal lines Indicate a hypod e la _
secondary storm‘sssumed to have occurre hover
the basin on March 29, I9I3, In the posntion s own.
The computed hydroqraph, resulting from the
combined rainfall is shown on dwq.w-PP-O-IA63

Scale lo 0 I0 20 Miles

 

TRANSPOSED POSITION
UPPER TENN. RIVER BASIN
STORM OF MARCH 22-27, I9l3

 

FLOOD CONTROL INVESTIGATIONS

TENNISSEE VALLEY AUTHORITY
NAT... CONTROL PLANNING DCPAR‘I’HINT

.UIMITTID It I APPNOVID
...-2‘ _ 'I—I
xuoxanLt. 9-zo—3s w P? o IC54R

 

__.___---e-——-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

__ _ T.“ ---. _. -.

 

 

700 ——~———.— , - . . . _ . -. 7, , , _.___,,+
600 - 2
500 . ..- u
U)
u. . . ‘ I I
u 400 . ~ . i . - was. a
o ,
O
0
Z
3 3
o 00
..J
LL
200
I00
0 f s i Y T 1
20 2 2 24 9
MARCH
NOTE:
Ohio River Valley storm of March .
' 22-27, I9I3 assumed applied to the
Upper Tennessee River Basin in position ESTIMATED HYDROGRAPH
shown on dwg No. w-pp-o-ICS4 together '
with a one-day hypothetical secondary TENNESSEE RIVER AT
storm assumed to occur two days after CHATTANOOGA TENN
the above storm and in the position '
_._l__1_.L__-L_F__ shown on the same drawing. TRANSPOSED I9I3 STORM
I f F 14 A 907. runoff factor was assumed and
, the crest reduced to a discharge in FLOOD CONTROL INVESTIGATIONS
m‘I°"'I"°'f"'+""T""' cfs equal to 5000 i/ Drainage Area (sq mile). 32:5:fififimtfil’.filrfifim,
”Mm com, The drainage area above Chattanooga “Hum. . o- ~o .mom
veco.D.J.f. is 2|.400 sq miles. “0&1 ' ’ ._ ._ __
cu-o.,.- __ .. ..fikflfl/ l KNOXVILLI s-so-seLvflpp OI IA63 R

 

 

 

 

 

 

—_g.l_l.,..l_l-n 56

 

 

. ‘ 1"“ . 3:: T ‘L \ 'fi ‘ .
. - I "G. ‘aa—gafliwa'f» Y J W? ‘ i'fl.
‘ , P ' ' .

. , I‘ I
\
.‘~ - .
. O "‘, a '1-
1'7 \ I.
s ' ' I
.r‘ ‘ L
"

 

NOTE:
The Isohyetel lines shown indicate the
position of the Ohio River Valley storm
// of Janus? I2-25, l937 as transposed
over the pper Tennessee Basin in such
a location as to produce a maximum
crest at Chattanoogeflenn. ~
The computed hydrograph resulting
from this rainfall is shown on deg.
w-rr-o- IA54. -

 

Scale l0 0 lo 20 Miles

 

 

TRANSPOSED POSITION
UPPER TENN. RIVER BASIN
_ STORM or JANUARY I2-25. I937

 

 

  

  
 

 

 

 

...“ ' F L023." tCiONTROL INVESTIGATIONS

'1 - 8t ‘

- ...... ......0. sacrum.

I: ""- 7 ”"7
- 2f ,2. «

  
 
 

 

 

 

 

 

 

 

   

“WW~ .. -

 

 

 

 

 

700 ‘ i . -__-._,_4___, _E_.___._2Jj_ _

 

 

600 A V T J ' ' —o—~——o—_+——.—__+—- -..-fl

 

500

 

 

 

 

 

 

 

 

 

 

 

    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

CHI-9019

in , .

LI. I

U I

o 400 WALES, _c_.

O

O

| I

Lu

L9

9:“

J: 300

U

22

O

200
IOO
0 I I i T* T I T i Ii
20 22 24 26 28
JANUARY ' FEBRUARY
NOTE:
Ohio River Valley storm of January IZ- ESTIMATED HYDROGRAPH
25, I937 assumed applied to the Upper '
Tennessee River Basin in position shown TENNESSEE RIVER AT
on dwg No. w-Pv-o-ICSB.
A 90% runoff factor was assumed and CHATTANOOGA’ TENN
L I I I the crest reduced to a discharge in TRANSPOSED I937 STORM
1 I 1 I 1 cfs equal to 5000 VDrainage Area (sq mile).

v I .V The drainage area above Chattanooga FLOOD CONTROI; INVESTJSIYATIONS
1%?01'0 as 21,400 sq mu... .12::::.2§3.¢::I.:2.0.2.0....
DWJ!” COIMID IUIII‘H’IO I CON. NDCO Imouo

'm..p.2'.fi. ”'0 n wwaemyv .XJJ‘AH-
.....L 12479222110111- , I KNoxVILu: 9-29-39 WTPP]OT IA64 R .

 

EXHIBIT 33

 

MASS RUNOFF+MILLION DAY-SECOND-FEET

 

 

 

 

2 3 4 5 6 7 8
DAYS

NOTE:

The mass-duration curve for each flood was
obtained by summing, in order of magnitude. the
daily flows downward from the crest of the flood.
Mass-duration curves for the floods of l867, l875.
l886 and I9” are shown as determined from actual
flows. and from the actual daily flows in each flood
increased by the ratio of the maximum assumed
flood crest of 730,000 cfs to the actual crest.

 

l2 I3 I4

I5

 

MASS-DURATION CURVES
HISTORICAL AND HYPOTHETICAL
FLOODS - TENN. RIVER AT
CHATTANOOGA, TENN.

 

 

FLOOD CONTROL INVESTIGATIONS

Tlflfllllll VALLEY AUTHORITY
warn com nanmno avenues-r

 

non-mun.

42441.44 5213.13.44

 

 

 

Min—gt-

 

o-za—aa [ w [RP 01

IASIR

 

 

 

 

 

 

Jz“_-, _‘s- I - i ' _-_- _

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

F— -— 1 --.-__-.1-__ =—~ ..
800 w 7 . 4 w 4 T
. I I I y [T I i I I
I ’ l I l I l
I #I fit I I 4 i - I L L I g
I I ' I r I T— V l ' I T ”k R
. I I . Nun 730.000 cfs ;
. I i I ‘ I i I I
700 * T I I : f I I l I N I 1 T ‘ 7
I I / I I \ I ; I I l
I J I i I I I I I I
I V ' T l I Y i “
I ' : I ' I l I I I.
i i l ; ' I ‘ I
I ‘ 4 I I I '
I - I i -_ _-_I__
I z
«n ' I ' I
.._ 500 I .
u / I ;
o I I 1
8 I l i I T
_ I I
z 400 I I I
.. j i I
g A.
< i I
I 1 I I
5 300 I T
° I
I
I
200 41
/ x\
I/ \
0 4
2 4 6 8 I0 I2 I4 I6 I8 20 22 24
DAYS
NOTE:
It Is assumed that a flood of this magnitude HYDROGRAPH
may"1 o+ccur taro??? the flogd season MAXIMUM ASSUMED FLOOD
It ' I1 ’1 I5. '
:UbSIIIJ‘TIfafI.;Io:dOS a: :ormewt'gt 1:: magnltude TENNESSEE RIVER
I 1 I "“3! ”CU" '9 I”. “ AP"_' '5' AT CHATTANOOGA TENN
I r I L FLOOD CONTROL INVESTIGATIONS
.?:::::.£:.'A:::l.:9.rrf.:m.
' 4 a - 2 Mfim
KNOX s-zo-sa w PP o ””2310

 

 

FII 911—5:an

 

 

 

\ \ \i\mi %\

 

   
  

NUyOSV II.

I
l -
kigoo 68.4 3,3-551'1441;

 

 

 

 

 

 

Tennessee .. I \ ... Noam CAROLINA
GEORUA 'f "—‘f "—4-- ” " GEORGIA
LEGEND
. Stream Gaging Shim
R R mi new flood
- Mammm rates “In were
MAXIMUM RECD OED UN OFF RATES "a" Mm“ m " “U" a. a,
DRAINAGE PEAK FLOW DATE Tenneseee'and publiohd recorde 0‘ the u 3
STATION AREA C [3. PER O “his“

30. MILES 30. MIL! OCCURRENCS ‘7’" drairuga amend minim race-dd

. IQ” h .

Cherie-ha 2,!00 t! l 4 -3 '20 '9': are pelican Wemgw.‘.§3?qm .“
Bellanca l,IBO 415 "40-00
Mary», 4 l0 5“ 3 - i9 - n
Nye-villa :90 “a it 7 -2!
Twila IOO 755 l ‘ S ' 3
Raw 27! 359 l - 22 ' 02
Parhoville 500 20.4 7 - no -16
(ml 530 40‘ 7- lo-Ie
M‘Harga 45l 29 O l - Zl - £2
Dial I75 52.5 7 - 9 ‘ 06
Bin R' t 3| 60.! 7 ‘ 9 ' U

 

 

 

 

 

 

Maximum length of reord- ”para at Manly

Awe, la a! recent - leper-

Low rate a U‘Harg 4 due to stat-or being establvehed In My
l9l‘l and therefore no record is available for ma flood-

“I O 5 ”HI”

Lu fl

 

 

 

MAXIMUM RECORDED
RUN-OFF

 

 

 

 

353,, m ....o. cm HIWASSEE-OCOEE RIVER BASINS
' ' ‘ SE UTHORITY
9M. ”64 PR EPA R E 0 BY‘ mg“ i ‘ TWEAhTIERE EON TEROLVAPIC lA-IENYINGADI PART Ml NT

 

 

 

 

 

men-(J2, . gmwg%

M0 - §NOXVILLE Ia-la-silslcplsl BAG 0

"x

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. . . 4 9.4.... . - . -. . . e -
J—I . .a CH. A m. 9.. . .2 rr . r . . .
O¢o<n _ 0 Ti 3—3;".-. 54.36:: --.g ................ ....335
K .
.ngNNNx Vows SNVN -..WEQHQE . ammo to? m .6 9.68.. 3.3 9:3 momma» .528... .. uwofili
. A . . . . ..Wi‘-g ......... 1c. IOO
23...... 3.2.3.8... 8E3...- meo :m . 23:26 mm .6» no 63%.. 95 3682 umcgzaaancm £55 .8988 355.8 ......-
hzulhceauo 02.215.305.100 cure! SOL» mwma+m $0.5 . mco_+m+m 034. 9.: +0 mmaafi cm03+mn Encoiflmu 59750:: :8
mZO_H“%HW=W%HJWNHM.M2W”Wood of we $333 629:. of 60.: 30298 come no; 5.36686 35 Sa .NNQ I
..m Lanswumo 2 3m. ._ €262. Lo notoa of co» Banzai 2m . 68:3: 3:6 m.
. . £52841 t... £62m 4&qu oz; 9.: E 323 9:8 of +0 32.1868 3 ammo
ZZUF MUZ<_I_U1 ”—(UZ +533 2: 2 3668284 ncm 68:3: 3:6 3%: ammo o ca not 82, moon-a
muZm uuww< _I Jog. .8 63,533 2 ma. ._ 62:98 gee-1 new 22 ..m #5683 2 82 .2
>> a: 2 68“. .39.. Loco-Do 36... 2.33 o coma mm; 52; . $385... . mucmzoa . . .
cmwc xmtaw 2.65280 6.: of Co mama Emmmca of 2 2&0 E85 32:». or: A .33 qaam mmm; vm oMFOLconwsfiowww
mOOOJ... ...O ZOEmePmE L52
one 52 .50 :2 «.2 as; 3.1 2:.
mm m. n nm 2 n 3 2 mm m. m. 30. n 3 n. ammo. 8.... no
N.
m.
cm
em
em
5
w.
mm 9
3
m m.
an: own. o~2 22 com. W
g > - b h b p n b p p p h b h P n b h . Q B
3
H
N- Av t s.. c . . t N- W
. - 3
C C v v n 3 v
o. H a .4 I c 4 .w : c. 3
z A L.
°~ . 4 . 2
. 2..
4-
a a
a a
A.“ 2.8“". oz k I 9.8:. oz
«a Na 4

 

 

 

 

 

TENNESSEE RIVER

 

NOTE. '. ,
The isohyetais shown indicate the
- position of the East Tennessee storm
Of March 22-23,l929 as transposed over
-the Hiwassee River Basin in such a . Scale IO 0 IO 20 mm

location as to reduce maximum crests

at HiwaaSee am and at mouth of

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Hiwassee Rfivei;i d _ - . - *
The com u ed rOgraphs resulting '
from thiapramfalyar‘e Shown on dwg TRANSPOSED POSIT'ON
.S'GP'°‘5A‘2- 3‘ "'P""3"AZ°- STORM OF MARCH 22-23 1929
Due to lack of satisfactory Topographic
[ T Maps relief shown was prepared from HIWASSEE R'VER BASIN
I I 'Aerial Mosaics with control from
eem-..4..1.. M... ereeeeee eme- FL°°2.Rearmsizvséilaei'°~s
no 'I . . wane comm names permit-m
Olin IMO . ' . OUDIlTTID ~ Recofi‘TE—W
. -- - ...-24 ____________
cum”. -W KNOJVILLE 7-5-39 INIPPIl 13 2A 90‘

 

 

 

Funso -lo.o'2
runs-luuu * - -__

7 I. A -1. ———— —-— ‘vVn'j _ . ~ ——~

 

 

‘\

 

 

 

. q < oe
, u ‘
I O ... * ~+~—4—‘ o-.e..
' I

r 4'- 4—. ,.
I

v
vo—H
I

IOOO C.F.S.

 

IN

DISCHARGE

l8 24 30 36
TIME IN HOURS

NOTE? -
Eastern Tennessee storm of March 22-23, MAXIMUM ASSUMED FLOOD
l929, assumed applied to the Hiwassee .
River basin in positlon on dwg w-pp-i-322A8 HIWASSEE RIVER

A 90% runoff factor was assumed and

the crest reduced to discharge in ’cfs H|WA$$EE DAM N. C.
equal to GOOOWrainaqe Area (sq miles).

DFSIDSQC area at this DOIDI IS 977 sq miles. FLOOD CONTROL INVESTIGATIONS

TENNESSEE VALLEY AUTHORUTY
emu ruinous ovum-nu

-
O

 

. l ‘ .
............... _ (‘

----.éc/l ' Supersedes S‘GP'9'5A7” KNOXVILLE Io-io- 5 or o SAIZR

 

,FU l53-I0.0I
l"-

I
-

 

 

 

 

l532 5:532

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Jy-e—ov '

“ e‘- fi ,
».+o-+—.'l-—+—o—e ’ 1‘ ' ». v—e—‘oe o
e.,._.‘_._,_._§_r_‘ >— * < . 4,. : ‘er§FP.-eeelb+«7+v~ +9L9¢0He
romH—e—hq H y Y , r YT ‘fi,

'500 YY ‘CO—eo»>oe»o fro—Y 4 .H.« HCfVoO—O—o&< A...._.I.e.e HveH-..~y
Mme—ML H '47‘~~‘ «coveO—o—e-e >—-4— ‘09-—H—- ~~N4ooeHe kw e—erO—Hrkoin-‘too-GOV
H+‘H+‘_‘ -1 -..»ov Ho+e++vo T —+-e—H t-¢-?H-'t4—v-Fr+’—r—o "« 4++.,.~. ..h
‘Hefi—M H ""H ""‘ ‘+‘ H‘? ‘T‘Ttt'twr‘f? ' “WtHVH-5‘4 v y~oHH
‘--,,,-A,.__+_‘ l—4 ‘ .4. o , o o uo—v etqu er—ee‘“ QHT—n—Tawo-T—eea‘ MfH o+<r>e< #veHo tHY*
>H.§.o—o-e—e H ’H‘*4 'tt—trH-O—fs ‘ —e—¢—OH>~>e«+—+aoe ~o~09ee—vT-+—+ veHv4—o—v4

eo»-o~e< O‘Plb? eeaoeoeo—g—e—e—i ea—eeT—eeH. eson-o He,.et.-.e4.§-4 +4—QH‘O4v—l
yeeee.o—o—+—§_4 “* +>—‘—¢v—v~-9~ rr-veehv el-a xoe<twe$~ a—>+.e—->e.+ -+—oe—+o-9—+—i
H WW +4 ore ~++H v r<eo e— oqp e wAee e e o reeeer—q—a f -'H~¢ ...». ‘T-{PO eeo—o—o—o—ee 7%
ee-eeeeee e , e _ A» ,7 - u ~ l460

I460 V eT—e—efl 4 o H- eeoey +>1t~0H"—H‘ +1 eee e He~ .+e to». o H o oar—ve’q v o .+o-H»¢ e
.eeero—e-o-ow—ef—i O—r-s H »>— +~ Q —>+++e-.—. b—o—o - ¢—. . «e e e—o—e-o—e—o .H Lee- . . L. 4
M’TW s‘O-«vev e’O‘Hoo «4 e .kTQQ-fi 'H'HVO—*‘ >e-e -- o—fv 4-1«roHe
Hwy—9.4+ ‘Htrt 0 o-o‘erto—wvo—v—oo-e-gre > o—e—+ -.e>.¢—.~A.»r-e1oo~te+fi
:I—Q—fl-Q—‘ahO—F H tree—tr §+oHo ‘o kg: H .44 .H .. >--<.,4.._H
MH‘M‘ Wt—t—i ot—fe—Qt4-9 feeke—t—oo—l~e4»—+eof—¢— pas-4 o~e'O—H—4—eov—i—fooofvvo+1
Who—«sew ' ~04- _ ‘ Ttt—Ht-s-l t —+ Hv~o—-Yrt—O—'¢+ 44+. §.H< oe ~44—9ot—ooe—4
e-yo—T-fe “FM e ‘H‘tt . r—f ’t‘MI—T‘thtva Fe .H$¢ooe ,,, eeoeeef—e—le—i
WM . e . +9 #4 'v‘O—eo A ’t—vro If‘t—ft4H—l +¢~fT+e eu—‘eTfI—eeez-efo—er‘ e. e» b-v e—ero4veo—e—T—4 I4

.420 M—e’4-Q—I—94 kt—H H e— - A A 4 I—T— ' 7 V

v y V v v Y 7
THsg—freeo—e f. .44 o+eLo—e%:i:i+§vi—%Hrf t++~e ouf e04-6Iee—fre—lee—da
9 +9— Iee—eee-o—en >—' e th-oeho—t O—H-O—é—o VH‘ .,,,+ ‘
" ‘44 t-t ‘4 .3 t—H»«>—r—+ v+++ , ~< e+++++++¢ I V 9+. +$1 0» H4 —6—<
*‘m ~voeoron—r moo-LTee—t—wo—é-ofi e v» Gee-ie >eeeeveHQp—4
‘ ‘ T - _ y— t o e

 

 

 

b—p-e—e
v v 1 v 7 r r v i v
>O~T+fi -+e+94~1efH—oee + «r—l‘e—l—le .4
H
I
1

 

 

I IlJIII ll

 

 

 

 

— -4
F... $44 9 o—e—‘eeLoo—H e 4e.- e Awe—e .4—0—1 e—e—o .H—Q e . H-‘ee—4—oeo . - QAH+ ..HA—o—ffil
l A . A A - t I
bffl-ore‘aee» >O— 4-4—54— Y ' _ V . Y ' A v, V 7 ¢ : ,,
l A ll . L, , i .
‘ _Y 1 v v y y yy . ~¢ » H++e a e447+..o+oo—+_4>s.p—oe.ge—'e .QH.e.. ‘+‘ .AMHeye' HA,
H++ >—+ W4 ‘ «4—. .-L. e+o .

i t—O—o‘ v-o—o v4-- 0 rev 6 . >Oeon—Q oHe .4 < e<e¢eoe+$q ...—+-
9 My e—h. e—He. H—qbee—oa . 6—. eg—ibe rA-o—e—eHeoto—vo+ H—o e+4-' r—Lei +—f—q—>+¢—<

 

 

 

. t—T44—06—H69—o-o-4—ivymep—p4—‘y—ii ¢e~~o ...HoH4«¢Ho++..—L«
H-e4—H H9 9 e44 Mvo ~9—c QH?—?—t— —e—-q 9+”; H'f"“ v+H>6+6i4—H4¢—1
‘ o-o—oHfi» 94—s. e. . e— .04 . .+—ye++—e>-o eHeec—q—H 4e . .-.—h ‘4 . . e-.e.H
96H t-“§¢‘CQQO~OU*Q§+H H.~O—HOH" +0..—O—">O-—< .+'OO‘+§.
oo—o—o >4f0 - v—v—c o— oo—<ee-e-f—e—+H To—ee—eeH—e Hv .o—o—e ea kvO—Q—O A—e—r-e ea

‘4» reoov¢ 04 —‘ ~ r-o '

 

 

 

->O¢Y+ 0"0‘

vO—fikfioeo ..4‘?‘>49H909i‘+1>+—O‘+-+‘0““ oeeoeoaeoe—oeq—e—e—z—eeew

o o-e— o—e—o ?o ooe—pfie—Hee-‘Hee—os e-e -. . 7e V-O—V" e o .e >V‘-T—.T‘ eeee.

L ob—y—oO—Y—4 9' iy—e—oeH—fo—o—e—‘—o—lefiepee.-¢e a.4..->eveeeo—e—H—+—¢ee.e.1
O—Ht meme—H“ Wweeeeoeo1e4¢eu>veeee e—f—Q—eoo-H
Hve o+ f—Q—o M. e—n-feH—e 5‘ 1%. e . e ++<>o . o v-O o+y<>:It—I—Q_Q_._.*+—i

’ e w‘rf‘r‘ H—. ...—.4 H».H—._+ e—e—pfi . . .,_§_i»—4 A. e o—o+e<»H—$+¢—e—.4

‘ O O

 

W. S. ELEVATION-FEET ABOVE M.S.L.

ii

ii

i

i

i

i

i
w.$.ELEVATlON-FEET ABOVE M.S.L.

. . ...
ve-o»-<>« k+¢ro¢~<+<r>e .H4H—4—ie—a—eevooeea. Ho H O—H-‘ —y~4—.¢+4_¢ev. I300
1 A A l ,

 

e i—o—a . e-e 4 . H. 9 ..-—... e oe-v—ee k4. . .C... ‘49—4 v ‘44. o ..-. e o—e—e-4—o—e—,—o—e—
e_e——.. ++ L. '¢+O—o‘—o~ _,,. o—eeoe—eee~o “—e-QOH‘HHQO-Oeb<--OO 4»o-eeo —o
' ‘ Y ' I’

 

 

 

 

 

 

 

 

 

 

 

 

 

eo-e .Hee»._¢—' kfa - A on . e .—>v—oe >—-¢—o—4 u. e e 0 HH fieo-p—o eeo—e-oeiLa—o—Jweet H
e-o—e Hep—i ‘HOeO—O-O‘O o—meb—Q—e—e 5+. o . ¢4t—He‘-o—’..Ho—< g—H .
ovew—o-o— oer. OQ—ofiHt—OO—vm I+¢OO~¢-e—o.fikr,‘4—egt—Q~H»¢+H.g ‘ o.
e >*~<‘~~—to —>~¢¢-—4—oO+_—f— Jeéeo-AH-ce4>ew§iefo'o‘—.1 -eo—o—e—oeHe-e
‘ Y Hoo-H v». aeoveeeoo- eoeom >m¢+0~<§ H—{Atoafie >9 _
a H--O—O—~¢ Lew—re >H+++o—._H— >—+~‘-— Hey»; . .. . t . +e ~ e +oeo~¢414~4 4—0—7v~o—O.O e 4 l e .e. .e, i o— e ' 4+—+‘
. e—ee H -4- r». «eo—o—oe Mv-e +9—O—f—Q- i . ; ‘4—+ 774e. e v e- . . ~e—eeo4~+§ e—e_ 44+, HT‘ _._‘... ’14 k b, _ kt“.
'260 Oeo—vaOO-f—é- i-o—o—H—eo—o—f-r >o—e—H+~¢—e— A 1 A A e A .260

2 4 6 3 5000 6000 7000
. DISC HA

 

PACITY

:QUATIONS AND ASSUMED LOSSES:
”"LLWMN .3:
. g Q- 90%}, - seesaw“ ( RNbrudenell, Univ. of iowe)
‘ ”HTIIE? it? 2:“. I‘ “‘ "’ ‘ “TH“ ,. “
”1:33; ‘H gram j ’ Where
I540 ' ' w 7 ' ‘ ‘ Ct 3.80 This value determined end the com
1:17: HM ' : i ' Egg discharge checked by model teete In the

e *t-O-J

4

 

 

 

IVA. Hydraulic Laboratory.
L- Length of Crest - 224'
. i , . 0' Design Heed on Crest -26.5'
:WETI-‘égdfirt ' H-Actuel Heed on Crest

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

—i T:::L“+—H+w~m ,
as vrwfitbfltw. - A EGULATlNG counuuT ,
>O-HH f—T‘ ‘7 A v
- :35: flags , A Entrenoe Loos-0.05?-
) I530 H 5 Alt we. 4‘“ ‘ e e
m 7%“; [We ‘f’m ”TM ~ .1 T Friction Lou-W (Manning's Formula)
> bkn+4wl htAIIIlv I . "I .
H’ t“. L 'I'A :7 3A 1111 . 1
8 W7 " LeeLMCeL * i. ; .. Nozzle Lou fi-9_2!;_
Wfi" 4ere—v—e- > L '
m < PhlfiWO—Hfli . H 1' ‘
U . *t, Mn.
i- ’- :::3:f:j.4—4. .14. we: Whore
‘ l- A It 11
( U :WH—Hwfiot-Tt ; L‘f¥£ T f 'I%; 1 V' VCIOC'fym P.
3 "J r ‘71 ‘:Ti:fiiz Li: frIIT Vi- Velocity et Smell End of Nozzle (Die.-7'-l0')
O u. WWI i ,’4 1 sittterl n-QOIO
I ”ease. tj' “A i ‘T‘Tti'lfi
5 z .520 T” 7i ;V :A ¢ {1431; L- Length ofStrelght Pipe-l80.5'
I 0 agent ft? 1' e; 39.1%? D-Diemeter ofPipe-U-G'
;: :rrrr Wasp; p; ; in? ~ C.-o.sa(King-s Hendbookof' Hydrauliee,Pg56) t
< :LTTT‘_IF.,1:-UA'§;* 7T metre e
t a ti ‘44 ; 4: tyr11‘f 1‘ GTE: ‘
_‘ r I. Q: .1531? ‘ olid portion of tellweter retlng curve ie booed on actual gage reedinge; ;
u; of Y; If ashed portion is estimated by eree-meen depth method. .
' 4 "TA' if ' :I‘Il¥ ]
m. v:v:¢vIIj J V fitl Y?”
3 i; {I 7; A e e. "f
I5I0 _. ~ A Lgrifjgxfi' . . DAM
a4 7 egg: Leg: HYDRAULICS
. "u. 'ffieeiii"*
1: is ‘ £33,? "
I505 ‘7‘ e ‘ ‘3‘: VJ» , yam"... AREA. CAPACITY a.
A ' “if r?" Curve.

 

 

1 1 1 1 I 1 0 20 DISCHARGE CURVES
ui-l-t-I-I-‘I-I“ HIWASSEE PROJECT

.-.”--- NN v u I v

 

 

 

 

 

   
 

 

  

(Gill.

MONCERINC DEMON “MITHINT
mm ..---"’H‘ .I ° ’ ICCOUHCNOIO AP'IOVEO
I.” -
-

 

b ................. >—--—--------—-——-

IZ-iO-a‘ll s I c I 4] IOK2l4R

 

 

 

 

 

 

   

 

'

Q
J ‘_
~ ‘ '
3 'I

“On”
”June;

q
I

L!

r‘t
k).

I‘M;

’ .

Us .§,_

....

 

 

MICHIGAN STATE UNIVERSITY LIBRARIES
I
ll
l

l lllll Ill lllllllllll

3 1293 03196 3261