American Society of Civil Engineers.

Transactions of the American Society of Civil Engineers (Volume 81) online

. (page 104 of 167)
Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 104 of 167)
Font size
QR-code for this ebook


erally confined to the river bed.



1222 FLOODS AND FLOOD PREVENTION

The organic act creating the Mississippi River Commission stipu-
lated among other duties that such surveys, examinations, and inves-
tigations of the Mississippi River and its tributaries should be made
as might be deemed necessary for the purpose of formulating plans
for the prevention of destructive floods. As a result of this legislation
the Mississippi River Commission has been engaged for the last 35
years in collecting data relating to the flood control of the Mississippi
River. An accurate hydrographic survey has been made of the river
from its head-waters to its mouth, with a topographical survey cover-
ing practically the entire width of the valley down to the head of
the alluvial basin near the mouth of the Ohio, and extending to a
distance of about 1 mile upon either side, with occasional sections
across the entire alluvial valley, below the Ohio. Maps have been pre-
pared on a scale of 1 : 10 000 with contours 5 ft. apart. Discharge
stations have been established on the main river and also on some of
the larger tributaries where the flow is not affected by back-water,
and the discharge curves have been developed from a long series of
observations. In the earlier stages of the work, daily observations of
sediment and discharge were made at several stations for periods of
over a year. In later years, discharge measurements have been
confined to prescribed stages above the normal high water and below
the average low water, in order to trace any changes in regimen that
might occur. Numerous gauge stations have also been established
on the main river and its principal tributaries where the height of
the river is recorded at least daily, and observations have been made
to determine the slope, at both high and low water, and other hydraulic
factors.

By the River and Harbor Act of June 18th, 1880, a project was
adopted for the construction of forty-one reservoirs in Minnesota and
Wisconsin to collect the surplus water from the precipitation of winter,
spring, and early summer, and release it during low water so as to
benefit navigation on the Mississippi River. In 1887 the scope of
the project was reduced to include reservoirs at the head-waters of
the rivers in Minnesota, only, and there have been constructed six
reservoirs with a capacity of about 96 000 000 000 cu. ft. and having
a drainage area of 4 535 sq. miles.

These reservoirs, while built for the low-water regulation of the
river, were much larger than any others that had then been built by man
for the high- or low-water control of rivers.* Careful records have
been kept of the rainfall, the contents, and discharge of these reservoirs
since their construction; and they afford most valuable data as to
the flood control of a river by reservoirs, particularly as there is a

• Three basins on the Ottawa River now under construction will have a capacity
of 168 000 000 000 cu. ft.



FLOODS AND FLOOD PREVENTION 1233

different relation for each reservoir between its capacity and the
area of its water-shed.

The surveys and estimates for reservoirs in Wisconsin, under
the original project, afford further data for the discussion of flood
control of the tributaries of the Mississippi River in that State.

The California Debris Commission has carried on a series of inves-
tigations of the Sacramento River and its tributaries, vs^hich are useful
in determining the flood control of that drainage area. Important
studies on the Allegheny and Monongahela drainage area have been
made by the Pittsburgh Flood Commission. A large amount of
data on the Miami River has been accumulated, in connection with
the work of the Dayton Flood Prevention Committee. The Scioto
River has been made the subject of an extensive study at Columbus.
And several flood commissions have been appointed by Governmental
agencies, such as the Ohio River Flood Board of the U. S. Army
Engineers and the Commission of the States of Indiana and Pennsyl-
vania, though few of these have yet had sufficient opportunity to con-
tribute any extensive studies upon stream regulation. Some of the
minor valleys have been surveyed by cities to determine the proper
source of a water supply, and a few corporations have determined the
capacity of certain streams to develop water power; but, on the whole,
there is a great dearth of reliable data as to rainfall, run-off, and other
physical facts needed by the engineer for the proper solution of the
problem of flood control, and the acciimulation of the proper data will
require many years of careful observation. Even with proper data
available, the determination of the method of flood control, which is
to be applied to a particular case, will require careful consideration
and the greatest exercise of engineering skill. No method can be
devised that will be susceptible of universal application.

Discussion of Suggested Methods.

(a) Reforestation. — If reforestation is considered merely from a
commercial standpoint, the value to the country of reproducing our
timber is so great that it obtains general approval, particularly of
engineers who appreciate more than any other class the disadvantages
to which the country will be subjected by the destruction of its forests.

But some advocates of reforestation have claimed that the destruc-
tion of our forests has increased the height and frequency of floods
and diminished the discharge during low water, and have based their
claim for reforestation on the beneficial influence of forests in pre-
venting floods and improving the navigation of our rivers.

The relation of forest to stream flow has been extensively discussed
in this country, in papers of recent date, by the U. S. Geological
Survey; U. S. Weather Bureau; Flood Board of U. S. Engineers, on
the Ohio River; Pittsburgh Flood Commission; Norton, on the Rivers



1324 FLOODS AND FLOOD PREVENTION

of Michigan; Chittenden, before the American Society of Civil Engi-
neers; Moore; Burr, on Merrimac River; Mead, on Rivers of Wisconsin;
and the Second Report of the National Commission for the Conserva-
tion of the Waters of the United States; and others. In Europe, in
addition to the classic papers of Gustave Wex, in 1873 and 1879,
Keller, Lauda, and Oppokow have recently discussed conditions in
Germany, Austria, and Russia, and a paper by G. Fantoli, 1913, sums
up the status in the valley of the Po. An interesting discussion of the
subject will also be found in the Proceedings of the International
Navigation Congress in Milan, 1905.

In these discussions the greatest diversity of opinion has been ex-
pressed, and even the advocates of reforestation as a means of ilood
control fail to give any quantitative determination of the effects of
forests upon floods. For the engineer to utilize any agency practically,
he must know that it will produce a positive effect on which he can
rely and which he can measure. The data as to the effect of for-
estation, soil absorption, and kindred methods of flood control are not
at present susceptible of quantitative analysis.

Over a large portion of the United States freezing occurs during
winter, and a heavy precipitation may occur when the ground is
covered with an impenetrable layer of ice and much of the vegetation
has shed its leaves. Such conditions prevent the utilization of forest-
ation or ground storage. This was the principal cause of the great
flood of 1912 which devastated the Ohio and Mississippi Valleys.

(6) Reservoirs. — The equalizing effect of lake storage on the dis-
charge of the St. Lawrence and Oswego Rivers is frequently quoted
as illustrating efficient flood control by reservoirs. But in these cases
Nature has been much more lavish in the expenditure of its resources
than man can afford to be. About one-third of the water-shed of the
Great Lakes is devoted to reservoir purposes.

The permanent overflowing of land which can be devoted to agri-
culture is costly, and reservoir construction is most economical when
confined to mountainous valleys where the area of tillable land is small
when compared with the volume of water that can be stored. On
many streams this condition materially limits the applicability of
reservoirs as a means of flood control, particularly in thickly settled
districts.

It also frequently happens that a reservoir can be more profitably
utilized for other purposes, such as power development, irrigation, or
the water supply of a city. There is a popular delusion that the same
reservoir can be utilized simultaneously to reduce floods, increase the
low-water discharge of a stream, and increase the water-power that
can be developed therefrom, but ordinarily its utilization for auy one
of these purposes precludes its efficient use for either of the others.
If a reservoir is to be utilized to diminish floods, it is essential that



FLOODS AND FLOOD PREVENTION 1235

it be empty when a heavy rain occurs. If to increase the low-water
discharge of a river, it must be full when the rainy season ceases. To
insure the first condition, it is necessary to empty the reservoir after
every storm, to obtain space to store the discharge of the one following.
To insure the second, it is necessary to close the outlets of the reser-
voir at relatively low stages, and when it is once filled, not permit the
surplus water to escape until the river falls to the stage when the
stored water is required for navigation. Both of these processes in-
terfere with the development of water-power, whose effectiveness is
measured by the minimum discharge permitted. The cause of the
trouble is the irregularity of rainfall in different years, and the neces-
sity of being prepared for a minimum rainfall when considering
navigation interests and a maximum for controlling flood conditions.
In years of average rainfall, it may be practicable to utilize a reser-
voir for different purposes, but it is reiterated that average conditions
do not enter into the problem of flood control.

Your Committee, however, does not intend to condemn in toto the
utilization of reservoirs for more than one purpose. In fact, it believes
that the practical solution of the flood problem in some valleys will
be found in permitting corporations to build reservoirs in which a
portion of the stored water can be utilized to a limited extent for
power purposes and the remainder for flood prevention, but in such
cases one of the conditions of the franchise must be the construction
of reservoirs far in excess of the dimensions required for the develop-
ment of the proposed power, and a rigid supervision must be exercised
by the authority granting the franchise, not only to see that its re-
quirements are fulfilled in construction, but also in operation.

There are in some localities opportunities for arresting the flow of
a stream during the progress of a flood by temporary storage which
limits the river discharge to that which may be safely confined within
banks. This can be done by detention basins which differ from storage
reservoirs in that they do not require the abandonment of the sites
occupied for storage purposes. Flood control in such cases is only
necessary during exceptional rains, and the basin itself may there-
fore be devoted to its normal uses most of the time, and only an ease-
ment on the lands occupied is necessary.

By this means cities can be protected at a minimum inconvenience
to the rural community. The use of detention basins has been adopted
by the Miami Conservancy District for the purpose of controlling the
floods of the Miami Basin.

(c) Barriers. — Barriers erected across the valleys of tributary
streams have been utilized to retain detritus and prevent its deposition
in the main river. When conditions are favorable, this is a practical
method of control.



1236 FLOODS AND FLOOD PREV^EXTION

(d) Channel Enlargement and Cut-Offs. — A stream flowing in an
alluvial valley adjusts its channel to its discharge; when the discharge
increases, the area of its cross-section is increased, and when the dis-
charge diminishes the area of cross-section is reduced. In a river like
the Missouri or the Colorado, carrying a large amount of sediment,
these changes are rapid.

Where a channel is enlarged for the purpose of carrying extreme
floods, it should be so designed as to give a velocity at all stages, which
will prevent deposits. If not so designed, the periodic removal of
deposits by mechanical means will be necessary.

A cut-off or straightening of an alluvial channel affects its slope
which the stream attempts to regain, producing excessive caving of
banks. Both stream-enlargement and cut-offs, while reducing flood
heights immediately above them, increase the discharge capacity
locally and increase flood heights below them. In alluvial rivers such
work may seriously affect the regimen of the low-water channel, pro-
ducing shoals, both above and below them, and should not be attempted
when questions of navigation are involved or bank erosion may seriously
affect other interests.

In streams of stable banks, and beds carrying but little sediment,
such straightening may not be injurious.

(e) Levees. — Every alluvial stream flowing in a valley which has
been formed by its own deposits, builds up its banks to a certain
height which is generally known as the bank-full stage. Below that stage
the stream flows in a more or less clearly defined trough and, at higher
stages, it spreads over the alluvial valley. When an unleveed river
rises above the bank-full stage there is an abrupt change in its regimen.
The area of its cross-section increases very rapidly and ceasing to
follow the sinuosities which its low-water channel has created, its
currents tend to flow across points, materially changing the slope.
These changes in regimen affect the movement of sediment, causing
deposits at some localities and scour at others.

When the river is leveed this abrupt change in regimen is pre-
vented to a greater or less extent, dependent on the characteristics of
the river and the distance of the levee from the river bank. The
function of a levee, as far as river hydraulics are concerned, is to
extend the natural banks up to extreme flood stages and thus continue
or possibly intensify the conditions which exist at a bank-full stage,
i. e., wherever a stream tends to scour or fill at a bank-full stage, that
tendency may be somewhat intensified at extreme flood stages if the
stream be controlled by levees which have been properly located with
reference to its banks.

There is another feature in connection with levees which it
is believed has been generally overlooked, and that is the increase



FLOODS AND FLOOD PREVENTION 1327

of channel storage which they create, and the resultant reduc-
tion of flood height from this cause. Under present conditions
of levee alignment, between the mouths of the Ohio and Red Rivers,
the Mississippi River has a storage capacity of 1 365 000 000 000 cu. ft.
(about 30 000 000 acre-ft.).* When we consider this enormous reser-
voir effect, it is evident that flood control on the Mississippi River
below the Ohio cannot be realized without the use of adequate levees.

(/) Outlets. — Outlets are at best of doubtful utility in reducing
the flood heights of sedimentary streams, even where additional lines
of flow to the sea or lake into which the stream empties are practicable.
The frictional resistance of the river bed limits their influence to a
comparatively short distance above them. In a delta formation they
usually require leveeing to protect the adjacent country from overflow
and in a river carrying a high percentage of sediment, either the out-
let or the main stream will diminish its area of cross-section by fill
and thus reduce the discharge capacity and ultimately defeat the
purpose for which the outlet is constructed.

Three natural outlets from the Mississippi River to the Gulf of
Mexico may be cited as illustrating this principle. The Atchafalaya
River leaves the Mississippi at a point about 200 miles above New
Orleans and reaches tide-water at a distance of 100 miles from its
source, while the main river traverses a distance of 300 miles before
reaching the Gulf. Although having three times the slope of the main
river, the entrance to this outlet practically closed itself at low stages
by natural processes of deposit and, in order to maintain navigable
depths, annual dredging has become necessary, and if this were sus-
pended the closure would doubtless in time become effective.

The Jump, an outlet which leaves the right bank of the river about
10 miles above the Head of the Passes, was once a formidable outlet,
but is now of insignificant proportions; Cubits Gap on the left bank,
about 5 miles above the Head of the Passes, which was cut through to
the Gulf in 1862, at first enlarged very rapidly to a width of over i
mile with a maximum depth of 80 ft. near the entrance, but has now
become a shallow outlet which is being rapidly obliterated by deposits.

So far as the Mississippi River is concerned, whose regimen has
been closely studied for over half a century, the observed effects of the
natural outlets that have been described and the effects of crevasses
in levees carrying one-fourth of the flood volume of the river with but
a small reduction in gauge heights, all point to the conclusion that the
use of outlets in any form for the purpose of reducing flood heights
will generally fail to secure the desired results.

• The Roosevelt Dam in Arizona stores about 1 500 000 acre-ft.



1828 floods and flood prevention

Conclusions.

Your Committee therefore submits the following conclusions:

Reforestation. — The effects of forest growth in preventing erosion
on hillsides are sufficient to justify reforestation for that purpose, but
there has been no quantitative determination of its influence on stream
flow which woiJd justify its employment as a method of flood
prevention.

Reservoirs and Detention Basins.— At the head-waters of streams,
storage reservoirs and detention basins can be successfully employed
to reduce flood heights. Which method is preferable is dependent on
local conditions. Their efficiency, however, rapidly diminishes as the
distance from them increases.

Cut-Off s, Channel Enlargement, By -Pass, and Outlets. — In clear-
water streams, whose banks and beds are not subject to scour, these
agencies may be profitably employed. In streams carrying a large
amount of sediment, whose banks are readily eroded, great care should
be exercised in their employment, always bearing in mind the fact
that the best conditions on an alluvial stream follow its confinement
to a single channel.

Levees. — As you proceed down stream the influence of reservoirs
on flood prevention rapidly diminishes, and the influence of levees
correspondingly increases in importance as a method of flood protec-
tion. On the lower alluvial reaches of long rivers, such as the Missis-
sippi and Colorado, they afford the only sure means of flood control.

Your Committee believes that it can perform no greater service
to the Profession than to call attention to the paucity of the data
existing in reference to flood control and to the damage which may
result from river regulation legislation, either by the Nation or the
States, calling for definite projects which are not predicated on full
and thorough investigation.

While valuable results have been obtained by National departments
and bureaus and State and local commissions, with the limited means
at their disposal, such investigations have generally been subordinated
to other questions than flood control, and there is pressing need of
intensive study directed especially to this end. Moreover, these agencies
have worked on independent lines, and there has not been the necessary
co-ordination of their observations. Progress would be much more
rapid if each local investigation could be brought to follow some uni-
form method of procedure and to cover all of the points included in a
standard outline, which would permit the fullest utilization of the
data obtained.

Your Committee desires to urge the great importance of the early
establishment and unification of systematic rainfall, run-off, and flood
observations, covering the entire country in far greater detail than
has yet been attempted.



FLOODS AND FLOOD PBEVENTION 1229

While the co-operation of all agencies engaged in river regulation
is desirable, the question of flood control is becoming of such vital
importance to the country that it should not be subordinated to ques-
tions of navigation or power development. The disasters of 1913 in
the Ohio Valley — involving a property loss of $67 000 000, and mdre
than 361 lives — and that at Kansas City, in 1903 — involving a property
damage of at least $10 000 000 — occurred on non-navigable streams,
and on some of the navigable rivers the interests affected by floods
are of far greater importance than those connected with navigation.
If considered merely as an accessory to navigation, the limits of in-
vestigation of flood problems will be vastly curtailed, and their use-
fulness correspondingly diminished.

The members of your Committee have received reports from many
sources covering investigations in connection with stream flow, and
have been much gratified with the prompt, cordial responses to in-
quiries for data. In digesting this information it soon became apparent
that reliable records of past floods were too meager and covered too
brief a period to justify at this time a general report on the floods of
the country.

A voluminous report might be prepared by compiling and tabulating
areas of drainage basins, slopes of streams, gauge readings, oscillations
in stage, and other physical data which the Committee has gathered.
But this did not seem to be judicious, since only a small portion of our
members are interested in such information and those who are would
doubtless prefer to go to the official records for the more complete
data required in their special work. Moreover, the expense and labor
involved in preparing and printing such data would be disproportionate
to the value of the results.

C. McD. TowKSEND, CJiairman,

John A. Bensel,

T. G. Dabney,

J. B. LiPPINCOTT,

Daniel W. Mead,
j. a. ockerson,
Arthur T. Safford,
Charles Saville,
F. L. Sellew,
C. E. Grunsky.*

^Endorsement of the Committee Report by C. E. Grunsky.

In concurring with the conclusions of the Committee on Floods and
Flood Prevention, the undersigned desires to emphasize the fact that
he is in favor of cut-offs and the straightening of river channels when-
ever conditions permit. The cut-off is not undesirable, as might be
inferred from the language of the report, but desirable, and yet there



1230 FLOODS AXD FLOOD PKEVENTION

will be cases where bank protection may be too expensive to justify
recourse to the use of cut-offs that might otherwise be thought advis-
able.

He desires, too, to add that, after the flood flow has been reduced
by all practicaTjIe means, the aim should be:

a. — To make the stream carry as much of the flood flow as can be
-carried between levees of reasonable height.

h. — To allow water in excess of the maximum stream capacity to
-escape from the river under control at selected points.

c. — To convey the water which is allowed to escape from a river
during flood stages, to a re-entry into the river or to a suitable place of
outfall, under such control as will prevent undue inundation.

C. E. Grunsky.



MINORITY REPORT.

By Morris Knowles, M. Am. Soc. C. E.

The writer regrets that at the last moment he finds himself in a
position which requires that he submit a minority report. His dis-
appointment is especially keen, as he was and is in substantial agree-
ment with the report as originally drafted by the sub-committee. It
is only because some portions of the report now presented represent
so wide a departure therefrom and because of failure to secure such
reconsideration of these changes by a reconvened committee meeting
that he feels it incumbent upon himself to present these views for the
consideration of the Society.

His objections to the report as now submitted relate to some ex-
pressions scattered through the paper, in which only a single view is
presented on controversial subjects on which it must be recognized
that much more information can and should be obtained. Such expres-
sions, he believes, promote controversy rather than helpful discussion,
and discourage rather than encourage further investigation. It would



Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 104 of 167)