Hydrology for- Urban Land Planning -

A Guidebook on the Hydrologic

Effects of Urban Land Use

By Luna B. Leopold

GEOLOGICAL SURVEY CIRCULAR 554

Washington 1968
United States Department of the Interior
STEWART 1. UDALL, Secretary

Geological Survey
William T. Pecora, Director

Free on opplication to the U.S. Geological Survey, Washington, D.C. 20242

 CONTENTS

Pam
1
3
7
11
11
16
16
18

ILLUSTRATIONS

Page
FIGURES
1-8. Graphs showing:
1. Hypothetical unit hydrographs relating runoff to rainfall,
with definitions of significant parameters - _ _ _ _ _ - _ - - 3
2. Effect of urbanization on mean annual flood for a l-square-
mile drainage area _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ 6
3. Flood-frequency curve for partial-duration series, West
Branch Brandyine Creek a t Coatesville, Pa., based
on d a t a for 1942, 1944-51 _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ - - _ _ _ _ _ _ _
4. Regional flood-frequency d a t a for the Brandywine Creek
basin, Pennsylvania _ _ _ - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
5. Flood-frequency curves for a 1-square-mile basin i n
various states of urbanization . . . . . . . . . . . . . . . . . . . . .
6. Increase in number of flows per year equal to o r exceeding
original channel capacity (I-square-mile drainage
a r e a ) , as ratio to number of overbank flows before
urbanization, f o r different degrees of urbanization - _ - 10
7. Annual sediment production per square mile for urbanized
and natural areas _ - _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ 12
8. Relation of sediment yield and discharge for an urban and
a rural or unurbanized area _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ - _ 13

TABLE

Page
TABLE1. Increase in discharge as a result of urbanization in a 1-square-mile

I11
HYDROL.OGY FOR URBAN LAND PLANNING-A GUIDEBOOK ON THE HYDROLOGIC

EFFECTS OF URBAN LAND USE

By LUNAB. LEOPOLD

This circular attempts to summarize existing opportunities offered by the principal physi-
knowledge of the effects of urbanization on ographic characteristics of the area, especially
hydrologic factors. It also attempts to express the location of hillslopes, soils, and streams.
this knowledge in terms that the planner can The existing pattern of land use and the ac-
use t o test alternatives during t h e planning companying distribution of woods and agricul-
process. Because the available data used in ture are parameters which over a period of
this report are applied to a portion of the years may actually change, albeit slowly. Roads,
Brandywine Creek basin in Pennsylvania, this villages, industries, and other manmade fea-
can be considered as a report on t h e basic hy- tures are more or less permanent and exert
drologic conditions of the Brandywine Creek their greatest influence in their effect on
basin prior to the beginning of major urbani- further development, especially through land
zation. Because the available data a r e not yet values.
adequate, this report can be considered as a Of particular concern to the planner are
compilation of tentative suggestions in the those alternatives that affect the hydrologic
form of an explanatory, not a definitive, hand- functioning of the basins. To be interpreted
book. hydrologically, the details of the land-use pat-
The application of current knowledge of the tern must be expressed in terms of hydrologic
hydrologic effects of urbanization to the parameters which a r e affected by land use.
Brandywine should be viewed as a forecast of These parameters in turn become hydrologic
conditions which may be expected as urbaniza- variables by which the effects of alternative
tion pro&eds. By making such forecasts in planning patterns can be evaluated in hydro-
advance of aetual urban development, the logic terms.
methods can be tested, data can be extended, There are four interrelated but separable -'
and procedures improved as verification be- effects of land-use changes on the hydrology of
comes possible. a n area: changes in peak flow characteristics,
PLANNING PROCEDURES AND HYDROLOGIC changes in total runoff, changes in quality of
VARIABLES water, and changes in the hydrologic amenities.
A planning document presented to a com- The hydrologic amenities are what might be
called the appearance o r the impression which
munity for adoption must always be more sug-
the river, its channel and its valleys, leaves with
gestive than coercive. This is true not only
the observer. Of all land-use changes affecting
because the planner is unable t o foresee the
the hydrology of a n area, urbanization is by
innumerable complications of actual develop-
f a r the most forceful.
ment, but also because there are many detailed
alternatives which would accomplish generally Runoff, which spans the entire regimen of
similar results. The planner is particularly flow, can be measured by number and by char-
concerned with both the constraints and the acteristics of rise in streamflow. The many

1
rises in flow, along with concomitant sediment water recharge and decreased low flows. Thus,
loads, control the stability of the stream chan- increased imperviousness has the effect of in-
nel. The two principal factors governing flow creasing flood peaks during storm periods and
regimen a r e the percentage of area made im- decreasing low flows between storms.
pervious and the rate at which water is trans- The principal effect of land use on sediment
mitted across the land t o stream channels. The comes from t h e exposure of the soil t o storm
former is governed by the type of land use ; the runoff. This occurs mainly when bare ground
latter is governed by the density, size, and char- is exposed during construction. It is well
acteristics of tributary channels and thus by the known that sediment production is sensitive t o
provision of storm sewerage. Stream channels land slope. Sediment yield from urban areas
form in response to the regimen of flow of the tends t o be larger than in unurbanized areas
stream. Changes in the regimen of flow, even if there a r e only small and widely scat-
whether through land use or other changes, tered units of unprotected soil in the urban
cause adjustments in the stream channels to area. In aggregate, these scattered bare areas
accommodate the flows. are sufficient t o yield considerable sediment.
The volume of runoff is governed primarily A major effect of urbanization is the intro-
by infiltration characteristics and is related to duction of effluent from sewage disposal plants,
land slope and soil type as well as to the type and often the introduction of r a w sewage, into
of vegetative cover. It is thus directly related channels. Raw sewage obviously degrades
to the percentage of the area covered by roofs, water quality, but even treated effluent con-
streets, m d other impervious surfaces a t times tains dissolved minerals not extracted by sew-
of hydrograph rise during storms. age treatment. These minerals act as nutrients
A summary of some data on the percentage and promote algae and plankton growth in a
of land rendered impervious by different de- stream. This growth in t u r n alters the balance
grees of urbanization is presented by Lull and in the stream biota.
Sopper (1966). Antoine (1964) presents the Land use in all forms affects water quality.
following data, on the percentage of impervious Agricultural use results in a n increase of nu-
surface area in residential properties : trients in stream water both from the excretion
I d hizr of Impeiviotix
products of f a r m animals and from commer-
risidpwtinl ai p a x u r f a c r area cial fertilizers. A change from agricultural use
fS[l tti (iicrrtnt)
6,000 - _ ~ - ~- ~ 80 t o residential use, as in urbanization, tends to
6,000-15,000 ~ ~ 40 reduce these types of nutrients, but this tend-
15,000 ~~
~~ 25 ency is counteracted by the widely scattered
pollutants of the city, such as oil and gasoline
The percentage decreases markedly as size products, which are carried through the storm
of lot increases. Felton and Lull (1963) esti- sewers to the streams. The net result is gen-
mate in the Philadelphia area that 32 percent erally aq adverse effect on water quality. This
of the surface area is impervious on lots averag- effect can be measured by the balance and
ing 0.2 acre in size, whereas only 8 percent variety of organic life in the stream, by the
of the surface area is impervious on lots averag- quantities of dissolved material, and by the
ing 1.8 acres. bacterial level. Unfortunately data describing
As volume of runoff from a storm increases, quality actors in streams from urban versus
the size of flood peak also increases. Runoff unurbanized areas are particularly lacking.
volume also affects low flows because in any Finally, the amenity value of the hydrologic
series of storms the larger the percentage of di- environment is especially affected by three
rect runoff, the smaller the amount of water factors. The first factor is the stability of the
available for soil moisture replenishment and stream channel itself. A channel, which is
for ground-water storage. An increase in total gradually enlarged owing to increased floods
runoff from a given series of storms as a result caused by urbanization, tends to have unstable
of imperviousness results in decreased ground- and unvegetated banks, scoured or muddy

2
channel beds, and unusual debris accumula- game fish give way to less desirable species.
tions. These all tend to decrease the amenity Although lack of quantitative objective data on
value of a stream. the balance of stream biota is often a handicap
The second factor is the accumulation of arti- to any meaningful and complete evaluation of
facts of civilization in the channel and on the the effects of urbanization, qualitative observa-
flood plain : beer cans, oil drums, bits of lumber, tions tend to confirm these conclusions.
concrete, wire-the whole gamut of rubbish of AVAI1,ABILITY OF DATA AND THE TECHNIQUE
an urban area. Though this may not impor- OF ANALYSIS
tantly sffect the hydrologic function of the Basic hydrologic data on both peak flow and
channel, it becomes a detriment of what is here volume of runoff may be expressed in terms of
called the hydrologic amenity. the Characteristics of the unit hydrograph, that

I ?Lag time
is, the average time distribution graph of flow
from a unit or standard storm. The unit hy-
drograph shows the percentage of the total
storm runoff occurring in each successive unit
of time. The standard storm may be, for ex-
ample, a typical storm which produced 1 inch
of runoff (fig. 1). Such data a r e derived from
o f stream f l o w the study of individual storms and the asso-
ciated runoff graphs measured at gaging
Center o f m a s s stations.
One factor stating the relation between the
storm and the runoff is lag time. This is defined
as the time interval between the center of mass
TIME, IN HOURS-
of the storm precipitation and the center of
mass of the resultant hydrograph. Lag time is
a function of two basin parameters-the mean
after u r b a n i z a t i o n basin slope and the basin length. These factors
empirically correlate with lag time if expressed
in the fmrn of the basin ratio (basin length L
divided by the square root of the mean basin
gradient, s ) . This basin ratio is also related to
drainago area. As drainage area increases, the
basin length increases and the average value
of slope generally decreases. Thus, natural
basin characteristics can be translated into
flood-flow characteristics.

TIME, IN HOURS- Lag time may be materially altered by the

effects of urbanization on the basin surface.
Figure 1.-Hypothetical unit hydrographs relating Water runs off faster from streets and roofs
runoff to rainfall, with definitions of
than from natural vegetated areas. This tends
significant parameters.
to decrease the lag time. The construction of
The third factor is the change brought on artificial channels, especially storm sew,ers,
by the disruption of balance in the stream also decreases lag time. As the time required
biota. The addition of nutrients promotes the for a given amount of water to run off shortens,
growth of plankton and algae. A clear stream, the peak rate of runoff (flood peak) increases.
then, may change to one in which rocks are
covered with slime ; turbidity usually increases, In addition to the basin ratio and lag time,
and odors may develop. As a result of increased the regimen of a stream, however, can be de-
turbidity and reduced oxygen content desirable scribed in many other ways, including flood

3
frequency, flow duration, mean annual flood, basins, for partly sewered, and for completely
discharge a t bankfull stage, and frequency of sewered basins. As in most studies the diffi-
bankfull stage. This is evidenced in past studies culty comes in translating these descriptive
of the effects of urbanization on the hydrology terms to quantitative measures of urbanization.
of a n area. Many different techniques of re- From data presented by Carter, values were
lating rainfall to runoff have been used, along read for a basin ratio of 0.12 representing a
with various parameters to measure the degree 1-sq-mi area having a n estimated length of 1.2
of urbanization. In order to evaluate our pres- miles and a n average slope of 100 feet per mile.
ent knowledge, it is necessary to express the It was further assumed t h a t in Carters study,
results of these studies in some common de- partly sewered is equivalent to 50 percent
nominator. sewered and 20 percent impervious. These
conditions provide some of the data shown in
Most reports on hydrologic effects of urbani- table 1.
zation present the conclusions in a form which
is more useful t o the hydrologist than to the Table l.-Increase in discharge as a result of u r b a n -
urban planner. This circular will attempt t o i z a t i o n in a 1-square-mile area
interpret the hydrologic conclusions of these [Discharxe is mean annual flood; recurrence interval is 2.3 years.
Data are expressed a s ratio of discharge after urhanization to
studies in terms that a r e meaningful to the discharge under previous conditions. Data from James (1965) have
no suaerscriptl
planner. Two forms of presentation will be
used. Percentage of area
served by storm sewerage
Percentage of area made impervious
n 20 50 80
The first is a slight modification of a method
1.0 11.2 11.8 12.2
previously used by several investigators, es- 0 21.3 21.7 22.2
pecially D. G. Anderson (1968) and L. D. 1.3 1.6 2.0
James (1965). The percentage of a n area 31.9 1.8 2.2
sewered is plotted against the percentage of the
20 { 1.4 - -

area rendered impervious by urbanization ; 1.3 42.1 13.2 l4.7

isopleth lines (lines of equal value of the ratio)
on the graph show the ratio of peak discharge
under urbanized conditions to the peak dis-
charge under rural or unurbanized conditions.
50

80
i 1.6
12.8
53.7
62.0
1.6
1.9
2.0
-
2.5
-
-
2.5
-
34.2
-
3.2
Such a graph will be different for different 1.7 13.6 14.7 45.6
drainage area sizes and for different flow fre-
quencies.
100
{ 2.0
-
2.8
-
16.0
3.6
Anderson (1968). a Wiitala (1961).
Martens ( 1 9 6 6 ) . OEspey. Morgan. and Masch
The second method utilizes a relationship be- 3 Wilson ( 1 9 6 6 ) . (1066).
I Carter ( 1 0 6 1 ) .
tween the degree of urbanization and the fre-
quency at which the original channel capacity As an indication of the change in impervious
would be exceeded. area resulting from urbanization, Harris and
Rantz (1964) showed that a n area near Palto
Table 1 is a n interpretation and summary of Alto, Calif., changed from 5.7 percent t o 19.1
the effects of urbanization on peak discharges percent impervious in a 10-year period.
based on previous studies. Results of the
studies were interpreted and extrapolated t o a One of the most complete analyses of urban-
common denominator of 1 sq mi (square mile), ization effects was made by D. G. Anderson
a practical unit of size for planning. (1968) in his study of the urbanization in
Fairfax County, Va., near the metropolitan
Carter (1961) developed a technique that complex of t h e Nations Capital. Andersons
followed the reasoning previously used by analysis follows the procedure suggested earlier
Snyder (1938) and that showed lag time as a by Carter, but Anderson included a larger
function of basin characteristics. F o r 20 array of data from 64 gaging stations. Ander-
streams in the vicinity of Washington, D.C., son closely confirmed the conclusions of Carter,
Carter developed this relation for natural but he carried the analysis further in a plot of

4
the ratio of peak discharge to the mean annual Espey, Morgan, and Masch (1966) analyzed
flood for different percentages of basin im- runoff data from urban and rural areas in
perviousness and for flood flows exceeding the Texas. To utilize this study, data were used
mean annual flood. For table 1, data from corresponding to a basin length of 5,500 feet
Andersons study were read directly from his and a slope of 0.02. I t was also assumed from
graph at the 2.33-year recurrence interval and his description of the area that urban could
expressed two separate conditions of sewerage. be expressed as 50 percent sewered and 20
The first condition was expressed as main percent impervious.
channels natural, upstream drainage sewered ;
James (1965) analyzed runoff data from a
this was assumed to be 50 percent sewered. The
44-sq-mi basin south of Sacramento, Calif.,
second condition was expressed as completely
within which 12 sq mi had been urbanized.
sewered drainage basin ; this was assumed to
From the basic data on flow, he obtained em-
be 100 percent sewered.
pirical coefficients used to route a series of
Wiitala (1961) presented data on urbanized synthetic flows by using a mathematical model
versus rural conditions f o r a medium-sized expressed as a digital computer program. The
watershed in Michigan. His data were trans- results were plotted in a series of curves which
lated into a ratio of peak discharges and it was separated the effects of flood frequency, drain-
assumed from his report that the urbanized age area, and degree of urbanization. Though
condition represented 20 percent impervious the derived curves do not present field data,
area and 50 percent sewered area. they also were incorporated into table 1.
Martens (1966) reported on three small Thus in table 1 are compiled, with certain
drainage basins in and near Charlotte, N.C. necessary assumptions, the data for seven pub-
Using flood-frequency curves from long-term lished and unpublished references which re-
records a t gaging stations in the State, he port measurements of the effect of urbaniza-
constructed a graph similar to that of Ander- tion 011 peak flow. Although -interpretations
son; that is, ratio to mean annual flood for were necessary to express the degree of
various degrees of basin imperviousness. As urbanization in quantitative terms, there is
before, the difficulty lies in ascertaining the considerable agreement among the data.
relation of Martens urbanized condition t o the
degree sewered. In reading from Martens
graph for recurrence interval 2.33 years, it is
assumed that the conditions he discussed in-
clude no sewerage and represent changes in
impervious area only.
Wilson (1966) presented data on flood fre-
quency for four drainage basins of 1.1 to 11.2
sq mi near Jackson, Miss. He presented his
analysis in the form of discharge of mean an-
nual flood plotted against drainage area size,
and he interpolated lines t o represent the per-
centage of the basin having storm sewers and
improved channels. I t is assumed that his
description 20 percent of basin with storm
sewers and improved channels would be
equivalent to 20 percent impervious and 20 0 20 40 60 80 100
P E R C E N T A G E OF A R E A I M P E R V I O U S
percent sewered. Similarly, his value of 80
percent was assumed to be 80 percent sewered Figure 2.-Effect of urbanization on mean annual
and 80 percent impervious. Rood for a 1-square-mile drainage area.
(Based on d a t a from table 1.)

5
 Data from table 1 have been transposed into catastrophic event is rare, and the storm of
the graph shown in figure 2. The ratios of peak ordinary magnitude is frequent. These events
discharge of urbanized to rural areas a r e pre- can be arranged in order of magnitude and
sented for different percentages of sewerage counted. For example, all discharge events ex-
and impervious a r e a ; lines of equal values of ceeding 400 cfs (cubic feet per second) can be
the ratio are drawn through the data. Briefly, tabulated from the record at a stream-gaging
these data show that for unsewered areas the station and arranged in order of magnitude;
differences between 0 and 100 percent impervi- the val?xes in the array can be plotted as a
ous will increase peak discharge on the average discharge-frequency curve. This has been done
2.5 times. For areas that are 100 percent for the gaging station on West Branch Brandy-
sewered, peak discharge for 0 percent im- wine Creek at Coatesville, Pa., for 9 years of
pervious will be about 1.7 times the mean record (fig. 3 ) . The theory and practice of
annual food and the ratio increases to about constructing such flow-frequency curves is well
eight for 100 percent impervious areas. Figure known. The plotting position or frequency
2, then, reduces the basic data to the same units often used is defined as
applicable to a 1-sq-mi drainage basin and to R=- n + l
the mean annual. flood. rn
A basin produces big flows from large and where R is the recurrence interval in years, n
intense storms and smaller flows from less in- is number of years of record, and m is the rank
tense but more frequent storms. The great or of the individual event in the array.
10,000

n
z
0
U
w
v,
ai
Lu
CL

I-
Lu
w
LL

2 5000
m
=>
U
z
- 1

w
. Q /

c3
ai
/
a /
I /
U /
-
v,
n

0 '
0.1 0.2 0.3 0.5 1 2 3 5 10 20
RECURRENCE INTERVAL, I N Y E A R S
Figure 3.-Flood-frequency curve for partial-duration series, West Branch Brandywine Creek a t Coatesville, Pa.,
based on data f o r 1942, 1944-51.

6
 Note in figure 3 that the largest flow in the 1 sq mi. The average annual flood is read from
%year record was nearly 10,000 cfs. The num- the upper graph of figure 4 as 75 cfs, and the
ber 50 printed on the graph means that there lower graph is used to construct the frequency
were 50 flows equal to or exceeding 500 cfs. curve in figure 5 pertaining to a 1-sq-mi basin
Once a year, on the average, a discharge value marked unurbanized.
of about 900 cfs will be equalled o r exceeded. The arithmetic for the construction of the
A slightly different result would be obtained curve is as follows:
if, instead of using the peak flow for each
storm, only the largest flow in each year were Recurrence
intervnl of
&ti0 to
inran ;unnu:il Ilischsuw
Recurrence
intrrval duration
included in the array. The principle involved nnnual flood flood 2 (cfs) series I
(years) (years)
is similar. The arithmetic mean of the peak
flows f o r the nine annual events is the average 1.1 0.55 41 0.4
annual flood. The statistics of this array are 1.5 .75 56 .92
2.0 .90 68 1.45
such that the recurrence interval of this average 2.3 1.0 75 1.78
annual llood is the same regardless of the length 5 1.45 110 4.5
10 1.9 145 9.5
of record, which specifically is 2.3 years. That
is to say, a flood of that magnitude can be ex-
Only the hixhest floud each yesir.
F r o m fisurc 4H.
: Obtained by multirdying ratios by 75 cls from fixure 4 r l f u r R
pected to be equaled or exceeded on a n average drninaae arc% of 1 sci mi.
All peaks during the ye;ii. T h c values in this column are mathr-
of once in 2.3 years, or 10 times in 23 years. matieally rchited to those i n thc first.

Studies of river channels have shown that The graph marked unurbanized in figure
rivers construct and maintain channels which 5 is constructed on semilogarithmic paper from
will carry without overflow a discharge some- the data listed in the third and fourth columns
what smaller than the average annual flood. In of the preceding table. The ordinate is the dis-
fact the recurrence interval of the bankfull charge, and the lower abscissa is the recurrence
stage in most rivers is a flow having a recur- interval in the duration series. An auxiliary
rence interval of about 1.5 to 2 years. scale gives the average number of floods in a
Urbanization tends to increase the flood po- 10-year period (calculated as 10 years divided
tential from a given basin. The channel then by the recurrence interval). Thus, the flow
will receive flows which exceed i t s capacity not expected to occur once in 10 years would be
just once in 1.5 to 2 years on the average but about 145 cfs and the fifth largest would be 75
more often. It is now proposed to estimate how cfs. The latter would also be the average value
much more often and to indicate the effect of of the largest flows each year during the 10-
this increased frequency on the channel itself. year record and thus would be the average
annual flood. It would plot, therefore, at a n
EFFECT OF URBANIZATION ON INCREASING abscissa position approximately at 2.3-year re-
FREQUENCY OF OVERBANK FLOW currence interval.
Taking the East Branch of Brandywine The effect of urbanization on the average
Creek a s an example, the flow-frequency curve annual flood is shown in figure 2, which shows
can be constructed for a typical subbasin hav- the increase in average annual flood for differ-
ing a 1-sq-mi drainage area. Figure 4A shows ent degrees of urbanization as measured by the
the relation of average annual flood to drainage increase in percentages of impervious area and ,
area, and figure 4B shows the flood-frequency area served by storm sewers. For convenience b

curve for annual peaks for basins in the these a r e tabulated as follows:
Brandywine area. The diagrams shown in fig-
Percentage Percentage Ratio to
ure 4 are similar t o those published in the of area of area average
sewered inznervmts annual flood
nationwide series of flood reports, U.S. Geolog- 0 0 1
ical Survey Water-Supply Papers 1671-1689. 20 20 1.5
40 40 2.3
From these curves a discharge-frequency re- 50 50 2.7
80 60 4.2
lationship is developed for a drainage area of 100 60 4.4

7
 RATIO TO AVERAGE A N N U A L FLOOD
0 - - N N
vr 0 cn 0 cn A V E R A G E A N N U A L FLOOD, I N CUBIC F E E T PER S E C O N D
4 VI
-
0
cn
0
vr 0 0 0 0
0 0 0 0 0

W
rn
n
c
W
-
W
rn
z
n
rn
-
z
-I
rn
W
< b
>
7
-
z
4
rn
>
W
v,
 DISCHARGE, IN CUBIC FEET PER SECOND
d d N h) w
ul 0 cn 0 cn 0
c" 0 0 0 0
0 0 0
I 0
? ; \ .
0 I
a

0
E
1 0
5

o
m
>
70
vr
5
P,

v1

0
 The average annual flood of 75 cfs was then curves representing urbanized conditions there-
multiplied by these ratios and plotted as shown fore converge at low flow values.
in figure 5 at the 2.3-year interval. These
Obviously the frequency curves in figure 5
values form the basis of a series of frequency
a r e extrapolations based on minimal data and
curves f o r combinations of sewered area and
require corroboration or revision as additional
impervious area. The shapes of the curves a r e
field data become available.
guided by the principle that the most infrequent
floods occur under conditions that are not ap- The flood-frequency curve under original
preciably affected by imperviousness of the (unurbanized) condition passes through a value
basin. of 67 cfs at a recurrence interval of 1.5 years.
At bankfull condition natural channels gener-
The most frequent flows are therefore in- ally can carry the flow having t h a t recurrence
creased by smaller ratios than would be the interval. If one assumes that this flow approxi-
average annual flood. Also, the most frequent mates the capacity of the natural channels, the
flows a r e decreased in number because low intersection of the estimated curves for differ-
flows from an urbanized area a r e not sustained ent degrees of urbanization with the discharge
by ground water as in a natural basin. The value of 67 cfs can be used to estimate the in-

Percentage sewered 20 40 50 80 100

Percentage impervious 20 40 50 60 60
I I
NONE MEASURE O F U R B A N I Z A T I O N COM P L E TE
Figure 6.-Increase in number of flows per year equal t o or exceeding original channel capacity (l-square-
mile drainage area), as ratio to number of overbank flows before urbanization, for different degrees of
urbanization. (Derived from figure 5.)
10
crease in number of flows equal to o r exceeding 3. Check dams, ungated, built in headwater
natural channel capacity. An auxiliary scale is swales.
shown at the top of figure 5 t o facilitate this. 4. Storage volumes in basements of large
For example, under natural conditions it is buildings receiving water from roofs o r
gutters and emptying into natural
expected that a 10-year record would show
streams or swales.
about seven flows equal to or exceeding 67 cfs,
o r channel capacity. But if the average annual 5. Off channel storage volumes such as arti-
flood were increased 1.5 times (from 75 to 112 ficial ponds, fountains, or tanks.
cfs) corresponding to 20 percent sewered and 6, Small reservoirs in stream channels such as
20 percent impervious, the new frequency curve those built for f a r m ponds.
indicates that 14 flows of 67 cfs o r greater Various types of storage volumes could be
would occur in a 10-year period, or a twofold used simultaneously in various mixes. The
increase in number of flows. Similarly, the effectiveness depends on the volume of storage
ratio of number of flows exceeding bankfull relative to the volume of inflow during a storm
capacity was read from the intersection of the peak period. Design criteria to guide city engi-
other curves in figure 5 with the ordinate value neers and developers a r e needed.
of 67 cfs to obtain the ratios plotted in figure 6.
SEDIMENT PRODUCTION
Figure 6 shows that with a n area 50 percent
sewered and 50 percent impervious, for ex- The basic data available for analyzing the
ample, the number of flows equal t o or exceeding effect of urbanization on sediment yield, though
bankfull channel capacity would, over a period sparse, have been summarized to some extent
of years, be increased nearly fourfold. in the literature. Especially valuable is the re-
port by Wolman (1964) who summarized not
LOCAL STORAGE TO COMPENSATE FOR PEAK only the data obtained from sediment sampling
FLOW INCREASE
stations in streams in Eastern United States
Urbanization tends to increase both the flood 2 but also studied the sediment yield from build-
volume and the flood peak. B u t the increase ing construction activities. Sediment yields '

can be compensated so that the discharge from urbanized or developing areas ranged
through channels downstream is maintained to from 1000 to more than 100,000 tons per square
any degree desired within the range which ex- mile per year.
isted prior to urbanization. It is obvious that
It should be recognized that sediment yield
reservoir storage is installed on a river in order
per square mile decreases with increasing
t o reduce the magnitude of peak discharge by
drainage area, but nevertheless it is apparent
spreading the flow over a longer time period.
that unurbanized drainage basins yield 200 to
Channels themselves provide temporary storage b'
500 tons per square mile per year, on the aver-
and act as if they were small reservoirs. Over-
age. These figures a r e slightly higher f o r the
bank flooding on to the flat flood plain is a way
farmed Piedmont lands, which may be expected
that natural rivers provide for temporary stor-
to produre sediment yield of 500 tons per square
age and thus decrease flood peaks downstream.
mile per year, such as the Watts Branch basin
This effect of storage has been fully investi-
near Rockville, Md.
gated and described (for example see Leopold
and Maddock, 1954, especially p. 36-49). The data on urbanized areas studied by
Wolman a r e plotted in figure 7 together with
The provision of flood storage upstream,
data from suspended load sampling stations of
then, will decrease flood peaks and compensate the U.S. Geological Survey as summarized by
for the increase caused by urbanization. This
Wark and Keller (1963).
storage could take many forms including the
following : In the graph (fig. 7 ) three bands or zones a r e
1. Drop inlet boxes at street gutter inlets. labeled A , C, and UC. Wolman and Schick
2. Street-side swales instead of paved gutters (1967) differentiated the following types of
and curbs. activity : A , agricultural or natural ; C, under-

11
 200,000
0

100.000
E
<
w
>
CL
w
a
w
d
-
zi
w
E
10,000
0
cn
E
w
a
Ln
z
E
z

.
z
0
F 1000
U
3
n
0
E
p. EXPLANATION Watts Branch
+
z
Ly
0 I I Y
X
X

U r b a n i z e d ( d a t a f r o m W o l m a n , 1964)
I
n
w
X
cn M o s t l y n a t u r a l ( U . S. G e o l o g i c a l S u r v e y d a t a ) X

100
I I

50
0.01 0.1 1 IO * 100

D R A I N A G E A R E A , IN S Q U A R E MILES

Figure 7.-Annual sediment production per square mile for urbailized and natural areas. Zones: A , agricul-
tural; C , under construction ; UC, under construction and undiluted.

going building construction, but highly diluted per year. Thus, Wolman and Schick drew the
before reaching channels ; and UC, undiluted distinction between agricultural, construction,
sediment yields delivered to stream channels and construction-undiluted.
from construction sites. For very small areas, Wolman (1964) said,
They found that when building sites a r e de- Because construction denudes the natural
nuded or construction, excavations are made, cover and exposes the soil beneath, the tonnage
and dirt is piled without cover or protection of sediment derived by erosion from a n acre of
near the site, the sediment movement in a rill ground under construction in developments and
or stream channel is very large in terms of tons highways may exceed 20,000 to 40,000 times the
per year immediately downhill from the con- amount eroded from farms and woodlands in a n
struction site. If the channel contains little equivalent period of time.
water except during storms (an ephemeral
stream), there is no chance for dilution ; during Figure 7 shows the data as a relation between
storm flow the sediment movement is great. If annual sediment yield per square mile and
the construction debris gets into perennial drainage basin size. The usual suspended load
channels, or for other reasons is distributed station is on a basin of more than 10 sq mi in
along a channel o r dispersed over a wide area, area. Seldom is urbanization complete for
the dilution lowers the yield per square mile basins of this size. ~

12
 EXPLANATION
Little Falls Branch near
k8

10,000 Bethesda, Md. (urban)

-X W a t t s B r a n c h near
Rockville, Md. (rural)

000

100

10

x
0 4
10 100 1000 10,000
D I S C H A R G E , IN C U B I C F E E T P E R S E C O N D
Figure 8.-Relation of sediment yield and discharge for an urban and a rural or unurbanized area.

13
 The data measured o r estimated by Wolman Keller (1962) compared t h e sediment rating
(1964) in small urbanizing, developed o r in- curves for Northwest Branch of Anacostia
dustrial areas show clearly t h a t the sediment River near Colesville, Md., a relatively un-
yield is larger by 10 to 100 times that of rural urbanized basin, and the Anacostia River basin
areas. Guy and Ferguson (1962) observed a n near Hyattsville, Md., which is partly urban-
increase of 250 times in an area near Kensing- ized. He found the sediment production t o be
ton, Md. about four times greater in t h e urbanized area.
To illustrate the difference in sediment Most sediment carried by a stream is moved
samples obtained during storm flow, actual by high flows. In Brandywine Creek, f o r ex-
data f o r two stations a r e shown in figure 8. ample, about 54 percent of the total sediment
The sediment rating curve, which is a plot of transported annually by the river (drainage
the discharge a t any moment in time against area 312 sq m i ) is carried by flows t h a t occur,
the concurrent rate of sediment transport, gives on the average, about 3 days each year.
a n indication of the order of magnitude of the
In the tabulation below, a comparison is
increase in sediment production from devel-
made between sediment yield from Watts
oped, as against rural, areas. The sediment Branch, a rural landscape, and Little Falls
rating curves in figure 8 a r e for stations near Branch, an urban one. These basins are of the
Washington, D.C. Watts Branch drains a n
size and type represented in East Branch
area prjmarily used for farming though urban
BrandyTvine Creek.
influences have recently extended into the
Drainaoe Tone per T o n s per
basin. Little Falls Branch near Bethesda drains area year year per
a nearly compeltely urbanized community, con- Watts Branch in (8s mi) 8q mi

sisting af Bethesda and parts of Chevy Chase, Rockville,

Md. Md. (rural) - _ _ _ _ _ _ ~ _ 3.7 1,910 516
Little Falls Branch
Note that the sediment rating curves tend t o near Bethesda,
converge at high discharges. One might sup- Md. (urban) ___ _ ~ ~ 4.1
_ 9,530 3,220
pose t h a t at those discharges the urbanized
areas a r e actually contributing no more sedi- Sediment production is importantly related
ment than the unurbanized ones. This is not to land slope. Using multiple correlation tech-
the case, however, owing to the fact that as a niques f o r a large variety of data from experi-
result of urbanization, the number of high flows mental watersheds, Musgrave (1947) developed
increases materially. Because most of the sedi- a multiple correlation in which the rate of
ment during the year is carried during periods erosion is found to be proportional to t h e 1.35
of high flow, the result is that urbanized areas power of land slope and t o the 0.35 power of
yield on the average larger sediment loads than the slope length. The same conclusion had been '

the unurbanized ones. derived theoretically by Horton (1945) and

verified by comparison with the percentage of
The difference in drainage basin size between area eroded in the Boise River basin, Idaho.
Watts Branch (3.7 sq mi) and Little Falls Sediment yield, therefore, is more highIy sensi-
Branch (4.1 sq mi) is not alone sufficient t o tive to land slope than to length of slope but is
explain the larger discharges in the latter basin. postively correlated with both.
For about the same number of sample storms,
note that Little Falls Branch data include dis- Some idea, however, can be obtained of the
charges varying from 20 to 1500 cfs. I n con- difficulty in keeping steep slopes stable after
trast, Watts Branch data (unurbanized) in- the original vegetation has been disturbed, par-
clude flows ranging from 7 to 150 cfs. A t least ticularly during construction. If, for example,
some of this difference is probably due to the land slopes of 5 and 10 percent are compared,
effect of urbanization on increasing peak flow the doubling of the slope would increase the
from a storm of given size, as discussed ear- erosion rate by 2.3 times.
lier. The two basins a r e only 10 miles apart Increased slope length does not have such a
and storms are comparable. large effect on erosion rate. Doubling slope

14
length would increase the erosion rate by only a t bankfull stage would probably have a veloc-
22 percent. ity of slightly less than 2.5 feet per second and
would be about 2 feet deep and 11 feet wide. In
Because a slope of 10 percent drops 10 feet figure 2, urbanization might cause a flow of this
in a 100-foot horizontal, temporary storage in frequency to increase 2.7 times, or 150 cfs. If
the form of depressions which might hold silt
this channel had to adjust itself to carry a flood
would be nearly absent. For land slopes above of 150 cfs a t bankfull stage, it is estimated that
10 percent, stream channels also would tend t o the new velocity would be about 2.5 feet per
be nearly devoid of areas or depressions which second, and the necessary depth and width
could hold up sediment during its passage would have changed respectively to about 3 feet
downhill. From a practical standpoint, there- and 20 feet. In other words, this stream would
fore, a figure of about 10 percent probably deepen about 50 percent and increase in width a
would be a physical and economic limit beyond little less than twice its original size. If such
which construction would be especially harmful erosion takes place through at least one-fourth
insofar as sediment production is concerned. mile of channel length in a drainage basin of 1
Any such limiting slope, however, would have sq mi, the amount of sediment produced by this
to be determined by detailed economic studies. erosion would be 50,000 cubic feet. A t 100
Wark and Keller (1963) related the average pounds per cubic foot, this amounts t o 2,500
annual sediment discharge in the Potomac tons.
River basin to percentage of forest cover and, This amount can be compared with the meac
separately, to the percentage of land in crops. annual sediment yield for Watts Branch, a n
Average annual sediment yield increased from unurbanized area near Rockville, Md. Annual
50 t o 400 tons per square mile per year, o r sediment yield of Watts Branch is 516 tons per
eightfold, as forest cover in the basin declined square mile. Thus, the channel erosion alone
from 80 percent t o 20 percent. Sediment yield under. the assumptions made would produce as
increased from 70 to 300 tons per square mile much sediment as 5 years usual production
per year, or fourfold, a s land in crops increased from a n unurbanized area of the same size.
from 10 to 50 percent. Therefore, one can visualize that as urbaniza-
tion proceeds, not only does construction ac-
EFFECT OF INCREASED PEAK FLOWS ON
tivity have the potential of increasing sediment
SEDIMENT YIELD
loads many thousands of times while construc-
It has been pointed out in the comparison of tion is in progress, but also the result of the
sediment rating curves for urban versus rural urbanization through its increase in peak flow
areas that the rating curves do not appear to would produce large amounts of sediment from
be as much different a s the values of sediment channel enlargement as well. This emphasizes ~

yield on an annual yield basis. It has been the need to provide temporary storage f a r up-
mentioned that a slight increase of sediment stream to counteract the tendency of urban-
concentration can make a large difference in ization to increase the number and size of
total annual sediment yield owing to the fact high flows.
t h a t urban areas produce a larger number of
WATER QUALITY
high flows. If the number of flows above bank-
full stage is increased by urbanization, the There is little doubt that as urbanization in-
banks and bed of a channel in erodible material creases, particularly from industrial use of
will not remain stable, but the channel will en- land and water, the quality of water decreases.
large through erosion. Computation indicates However, quantitative data to support this ob-
the seriousness of this factor. servation are sparse. There a r e two principal
effects of urbanization on water quality. First,
F o r example, assume that a channel is capa- the influx of waste materials tends to increase
ble of carrying 55 cfs at bankfull stage. I n the the dissolved-solids content and decrease the
Brandywine area this represents a channel dissolved-oxygen content. Second, as flood peaks
draining a basin slightly less t h a n 1 sq mi in increase as a result of the increased area of
area. The channel necessary t o carry 55 cfs imperviousness and decreased lag time, less

15
water is available for ground-water recharge. The soil, containing as it does a flourishing
The stream becomes flashier in that flood peaks fauna of micro-organisms, tends to destroy or
a r e higher and flows during nonstorm periods adsorb pathogenic bacteria. Effluent draining
a r e lower. from the seepage field of a septic tank tends
A recent study on the Passaic River at Little therefore to be cleansed of its pathogens.
Falls, N.J., by Anderson and Faust (1965) McGauhey and Krone (1954) showed t h a t the
provides quantihtive data on the effect of coliform count was reduced by three orders of
urbanization and industrialization on water magnitude in moving from a n injection well a
quality. Seventeen years of data for the flow distance of 50 feet through sand and gravel.
and chemical quality of the 760-sq-mi drainage In 100 feet the count was reduced to a small
basin were analyzed. During these 17 years, number. As f o r rate of movement, Mallmann
diversions of water for domestic and industrial and Mack (1961) showed that bacteria intro-
supplies increased more than 30 percent be- duced into a permeable soil by a septic-tank
tween 1950 and 1963. Returns of waste waters seepage field moved 10 feet in 2 days and 20
into the basin became as much as 10 percent of feet in 3 days and appeared in a well 30 feet
the water withdrawn. Analysis of the data away after 10 days.
showed that at relatively low discharge the Both the rate and effectiveness of the proc-
dissolved-solids content increased about 10 ppm ess of pathogen reduction depend on the type
(parts per million) between 1948 and 1955 but of soil as has been summarized by Olson
increased 75 ppm between 1955 and 1963. That (1964), who emphasized that position of the
is, during the period of greatest population ground-water table is a critical factor in the
growth the dissolved-solids content increased transmission of pollutants.
nearly 40 percent in a period of 8 years.
Studies by Wayman, Page, and Robertson
A long-term change in the average content of (1965) of the changes in primary sewage
dissolved oxygen was also noted. Between 1950 effluent through natural materials in conditions
and 1964 the dissolved-oxygen content dropped of saturated flow showed t h a t most soils re-
from a n average of 78 percent of saturation to moved over 90 percent of the bacteria from
62 percent of saturation. Further, the analysis sewage within a few feet of travel * * * [but
demonstrated that these average changes in there was] severe clogging in the finer-grained
water quality occurred in all seasons of the soils. They found, however, that dissolved
year. solids moved through the columns [of soil]
An aspect of population growth not generally virtually unaffected * * *.
appreciated is the large segment of population
using septic tanks for disposal of sewage. In The Same authors report on infiltration of
a given area this segment often becomes large polluted river water through sandy loam. ABS
before community water and sewerage systems [synthetic detergent] and coliform bacteria a r e
are built. For the planner it should be impor- significantly reduced by infiltration through the
tant to know how septic-tank installations can unsaturated zone ; dissolved solids do not seem
affect water quality in streams and in the to be removed * * *. Once a pollutant gets
ground. In the upper East Branch of Brandy- into the ground water (saturated flow) little
wine Creek, a basin of 37 sq mi, the population additional change in removal of ABS or dis-
in 1967 was 4,200. As of that date, there were solved solids, even for movement over extensive
no Community water or sewerage systems; all horizontal distances, is to be expected. This
the population was served by individual wells result is in agreement with the data * * * for
and septic tanks. Population projections indi- flow of sewage effluent through various soil
cate that the basin will have 14,000 persons by columns (saturated flow) .
the year 1990. During the initial part of this The data are not definitive regarding the
projected growth at Ieast, the number of wells minimum distance a septic-tank seepage field
and septic tanks can be expected to increase should be separated from a stream channel,
materially. but the application of data cited above with

16
general principles does indicate some tentative thermal pattern. The more solar energy a
rules of thumb which might be useful to the stream absorbs, the greater its temperature
planner. A perennial stream represents the variation diurnally as well as seasonally. By
intersection of the saturated zone (water table) greatly increasing the surface area exposed to
with the earth's surface. The observations in- the sun's radiation, the construction of ponds
dicate that, for soil cleansing to be effective, and lakes has profoundly affected stream tem-
contaminated water must move through unsatu- perature regimen. On Long Island, Pluhowski
rated soil a t least 100 feet. Owing to the gentle found that ponds having mean depth of about
gradient of the water table near the perennial 2 feet or less substantially increase downstream
stream and the fact that seepage water moves diurnal temperature fluctuations whereas ponds
vertically as well as toward a nearby channel, deeper than 2 feet exhibit a dampening effect
it would seem prudent that no septic tank on daily temperatures. For example, during the
should be as close to a channel as about 300 period October 31 to November 2, 1967, the
feet, if protection of the stream water quality mean daily range of temperatures at Swan
is to be achieved. The distance should prob- River, in south-central Long Island, varied
ably be greater from a perennial than from a n from 9 F in a reach immediately below a
ephemeral channel. (An ephemeral stream is shallow pond (mean depth, 0.5 foot) to 3 F
one which contains flowing water only in storm below Swan Lake (mean depth, 3 feet). In
periods.) In general, it might be advisable to J reaches unaffected by man's activities, the
have no source of pollution such as a seepage mean daily temperature fluctuation was about
field closer than 300 feet to a channel or 4F.
watercourse.
Under natural conditions, less than 5 percent
Even this minimum setback does not prevent of the streamflow on Long Island originates
the dissolved materials (nitrates, phosphates, as direct surface runoff. With the conversion
chlorides) from enriching the stream water of large areas of western Long Island from
and thus potentially encouraging the prolifera- farmland to suburban use during the last 20
tion of algae and otherwise creating a biotic years, the proportion of streamflow originating
imbalance. as surface runoff has increased sharply. As a
The only detailed study of the effect of urban- direct ons sequence, streams most affected by
ization on water temperature is that of E. J. street runoff may exhibit temperature patterns
Pluhowski (1968), some of whose results are that a r e markedly different from those observed
summarized here. He chose five streams on in streams flowing through natural settings.
Long Island for detailed analysis and found During the period August 25 to 27, 1967, a
that streams most affected by man's activities series of heavy rainstorms overspread Long
exhibit temperatures in summer from 10" to Island. Throughout this period, temperatures
15F above those in an unurbanized control. at each of the five observation sites on Con-
Connetqiiot River, the control stream, flows netquot River showed little day-to-day change.
through one of the few remaining undeveloped In contrast, temperatures in the upper reaches
areas of central Long Island. Temperatures in of East Meadow Brook, which drains highly
reaches most affected by ponding, realinement, urbanized central Nassau County, increased
or clear cutting of trees are significantly higher steadily in response to the relatively warm
in summer, but winter temperatures a r e 5" to
street runoff. Pluhowski found that by August
10F colder than those observed in reaches un-
27, water temperatures had risen 10" to 12F
affected by man.
above prestorm levels and were 1 5 F higher
Solar radiation is the predominant factor in than concurrent temperatures in the control
the energy balance determining a stream's stream.

17
 SELECTED REFERENCES Lull, H. W., and Sopper, W. E., 1966, Hydrologic effects
from urbanization of forested watersheds i n the
northeast : Upper Darby, Pa., Northeastern Forest
Anderson, D. G., 1968, Effects of urban development Expt. Sta., 24 p.
on floods in northern Virginia: U.S. Geol. Survey McGauhey, P. H., and Krone, R. B., 1954, Report on the
open-file rept., 39 p., 5 figs. investigation of travel of pollution : California
Anderson, Peter W., and Faust, Samuel D., 1965, State W a t e r Pollution Control Board Pub. 11,
Changes in quality of water in the Passaic River 218 p.
at Little Falls, New Jersey, as shown by long-term Mallmann, W. L., and Mack, W. N., 1961, Biological
data, in Geological Survey research 1965: U.S. contamination of ground water: Robert A. "aft
Geol. Survey Prof. Paper 525-D, p. D214-D218. Sanitary Eng. Center Tech. Rept. W61-5, p. 35-43.
Antoine, L. H., 1964, Drainage and best use of urban Martens, L. A., 1966, Flood inundation and effects of
land: Public Works [New York], v. 95, p. 88-90. urbanization in metropolitan Charlotte [North
Carolina]: U.S. Geol. Survey open-file rept., 54 p.
Carter, R. W., 1961, Magnitude and frequency of floods
in suburban areas, in Short papers in the geologic Musgrave, G. W., 1947, Quantitative evaluation of
and hydrologic sciences: U.S. Geol. Survey Prof. factors in water erosion-First approximation :
Paper 424-B, p. B9-Bll. Jour. Soil and Water Conserv., v. 2, no. 3, p. 133-
138.
Espey, W. H., Morgan, C. W., and Masch, F. D., 1966,
Olson, G. W., 1964, Application of soil survey t o prob-
Study of some effects of urbanization on storm
lems of health, sanitation, and engineering :
runoff from a small watershed: Texas Water
Cornel1 Univ. Agr. Expt. Sta. Mem. 387, 77 p.
Devel. Board Rept. 23, 109 p.
Pluhowski, E. J., 1968, Urbanization and its effect on
Felton, P. N., and Lull, H. W., 1963, Suburban hy- stream temperature: Baltimore, Md., Johns Hop-
drology can improve watershed conditions : Public kins Univ., Ph. D. dissert. (in preparation).
Works, v. 94, p. 93-94.
Snyder, F. F., 1938, Synthetic unit hydrographs: Am.
Guy, H. P., and Ferguson, G. E., 1962, Sediment in Geophys. Union Trans., v. 19, p. 447-454.
small reservoirs due t o urbanization: Am. SOC. Swenson, H. A., 1964, Sediment in streams: Jour. Soil
Civil Engineers Proc., H Y 2, p. 27-37. and W a t e r Conserv., v. 19, no. 6, p. 223-226.
Harris, E . E., and Rantz, S. E., 1964, Effect of urban Wark, J. W., and Keller, F. J., 1963, Preliminary study
growth on streamflow regimen of Permanente of sediment sources and transport in the Potomac
Creek, Santa Clara County, California: U.S. Geol. River Basin: Interstate Comm. on Potomac River
Survey Water-Supply Paper 1591-B, 18 p. Basin, Washington, D.C., Tech. Bull. 1963-11, 28 p.
Horton, R. E., 1945, Erosional development of streams Wayman, C., Page, H. L., and Robertson, J. B., 1965,
and their drainage basins, hydrophysical approach Behavior of surfactants and other detergent com-
to quantitative morphology : Geol. SOC. America ponents in water and soil-water environments :
Bull., v. 56, no. 3, p. 275-370. Federal Housing Adm. Tech. Studies Pub. 532,
136 D.
James, L. D., 1965, Using a computer to estimate the
effects of urban development on flood peaks: Water Wiitala, S. W., 1961, Some aspects of the effect of urban
Resources Research, v. 1, no. 2, p. 223-234. and suburban development upon runoff: U.S. Geol.
Survey open-file rept., 28 p.
Keller, F. J., 1962, The effect of urban growth on sedi-
Wilson, K. V., 1966, Flood frequency of streams in
ment discharge, Northwest Branch Anacostia River
Jackson, Mississippi : U.S. Geol. Survey open-file
basin, Maryland in Short papers in geology and
rept., 6 p.
hydrology: U.S. Geol. Survey Prof. Paper 450-C,
p. C129-C131. Wolman, M. G., 1964, Problems posed by sediment de-
rived from construction activities in Maryland-
Leopold, L. B., and Maddock, T., Jr., 1954, The flood Report to the Maryland W a t e r Pollution Control
control controversy: New York, The Ronald Press Commission : Annapolis, Md., 125 p.
Company, 275 p.
Wolman, M. G., and Schick, P. A., 1967, Effects of con-
Leopold, L. B., Wolman, M. G., and Miller, J. P., 1964, struction on fluvial sediment, urban and suburban
Fluvial processes in geomorphology : S a n F r a n - areas of Maryland : Water Resources Research,
cisco, Calif., W. H. Freeman and Co., 522 p. v. 3, no. 2, p. 451-462.

18
ir U.S. GOVERNMENT PRINTING OFFICI; ' 1968 0 - 298-448