Hydrology Lecture Note

TABLE OF CONTENTS
5 Hydrology of Un-gauged catchments................................................................................2
5.1 Synthetic Unit Hydrograph (UH)..............................................................................2
5.2 Snyder’s Method........................................................................................................3
5.3 Dimensionless Unit Hydrograph...............................................................................6
5.4 Extrapolation of flow data to Un-gauged Sites........................................................17

Hydraulic & Water Resources Engineering Department, AMIT, AMU

1
Hydrology Lecture Note

5 Hydrology of Un-gauged catchments

General
Due to different reasons that may pertain to technical or economical, there are some
catchments that are un-gauged. There should be a means so that the hydrology of
these catchments can be produced. Engineers and scientists have studied such
cases and came up with the acceptable means to the problem.

5.1 Synthetic Unit Hydrograph (UH)

Deriving a unit hydrograph for an un-gauged requires a relation between the
physical geometry of the area and the resulting hydrographs. Three approaches
have been used: formulas relating hydrograph features to obtain characteristics,
transportation of unit hydrographs, and storage routing. Basin characteristics
formulas usually pertain to time of peak, peak flow, and time base of the unit
hydrograph. When these features are established, the hydrograph can be sketched
to provide the necessary unit volume.

Synthetic unit hydrographs, once developed for a watershed area, can be used with
historical or design rainfalls to produce storm hydrographs at the outlet of the
watershed. As the watershed changes over time, the UH can be updated to better
represent land use and channel alterations. Synthesis UHs developed along two
main lines thought; one assumed that each watershed had a unique UH related to
specific watershed characteristic, and the second assumed that all UHs could be
represented by a single family of curves or a single equation. However, the formulas
all have certain limiting assumptions and should be applied to new areas with
extreme caution. Some calibration to adjacent watersheds where stream flow exists
should be attempted, if possible.

The first line of development was based on the rational method modified to include
the time-area curve for a particular watershed. Clark (1945) assumed that assumed
that watershed response would be given by routing the time-area curve through an
element of linear storage, which tends to attenuate and time-lag the hydrograph.
Each UH would be unique for a watershed, and this method thud represented a
significant improvement over the time-area method.

The second approach to UH development assumed mathematical representation for

the shape of the UH. A useful approach was advanced by the Soil Conservation
Service (SCS, 1964, 1986).

Hydraulic & Water Resources Engineering Department, AMIT, AMU

2
Hydrology Lecture Note

5.2 Snyder’s Method

Snyder(1938) was the first to develop a synthetic UH based on a study of watershed
in the Appalachian Highlands. In basins ranging from 10 to 10,000mi 2, Snyder’s
relations are:

t p  C t  LLc 
0.3
(5.1)

Where:
tp =basin lag(hr),
L =length of the main stream from the outlet to the divide (mi),
Lc =length along the main stream to a point nearest the watershed
centroid(mi),
Ct =coefficient usually ranging from 1.8 to 2.2 (C t has been found to vary
from 0.4 in mountainous area to 8.0 along the Gulf of Mexico),

Figure 5.1: Basin Characteristics

Q p  640C p A (5.2)
tp

Where:
Qp =peak discharge of the unit hydrograph (cfs),
A =drainage area (mi2),
Cp =storage coefficient ranging from 0.4 to 0.8 where the larger values of
Cp are associated with smaller values of Ct
tp
Tb  3  (5.3)
8

Where Tb is the time base of the hydrograph, in days. For small watersheds, Eq.
(5.3) should be replaced by multiplying t p by a value that varies from 3 to 5 as a
better estimate of Tb. Equations (5.1), (5.2), and (5.3) define points for a unit

Hydraulic & Water Resources Engineering Department, AMIT, AMU

3
Hydrology Lecture Note

hydrograph produced by an excess rainfall of duration D= t p/5.5. For other rainfall

excess duration D’, an adjusted formula for tp becomes:

t 'p  t p  0.25( D´ D ) (5.4)

Where t´p is the adjusted lag time (hr) for duration D´(hr). Once the three quantities
tp, Qp, and Tb are known, the UH can be sketched so that the area under the curve
represents 1.0in. of direct runoff from the watershed.

Snyder’s method is still one of the most popular methods because of its simplicity.
Caution should be used in applying Snyder’s method to s new area without first
deriving coefficients for gauged streams in general vicinity of the problem basin. The
coefficient Ct and Cp have been found to vary considerably from one region to
another.

Figure 5.2: Elements of synthetic unit hydrograph

Important relationships:
Basin lag tp

CtL and n are basin constants. (n= 0.38 and CtL = 1.715, 1.03, 0.50 for mountainous,
foot-hill and valley drainages of USA) Standard duration of effective rainfall, t r (in hours)

Hydraulic & Water Resources Engineering Department, AMIT, AMU

4
Hydrology Lecture Note

Peak discharge Qp (m3/s) of unit hydrograph of standard duration t r

Where A = km , Cp = regional constant

2

If a non-standard rainfall duration t R h is adopted, instead of the value t r to derive a unit

hydrograph the value of the basin lag is affected. The modified basin lag is given by:

Where t’p = basin lag in hours for an effective duration of t R. Therefore Qp,

Note that when tR = tr implies QP = Qps

The time base of unit hydrograph is given by Snyder as:

This equation gives reasonable estimates of time base for large catchments; it may give
excessively large values of time base for small catchments. Taylor and Schwartz
recommend

With tb taken as the next larger integer value divisible by t R i.e. tb is about five times the
time to peak. To assist in the sketching of unit hydrographs,

W=Cw.q-1.08

Where
Cw=1.22 for W75, and 2.14 for W50
W50 = width of unit hydrograph in hour at 50% peak discharge
W75 = width of unit hydrograph in hour at 75% peak discharge
q = Qp/A = peak discharge per unit catchment area in m /s/km3 2

Since the coefficients Ct and Cp vary from region to region, in practical applications it is
advisable that the value of these coefficients are determined from known unit
hydrographs of meteorologically homogeneous catchments and then used in the basin
under study. This way Snyder’s equations are of use in scaling the hydrograph
information from one catchment to another similar catchment.

Hydraulic & Water Resources Engineering Department, AMIT, AMU

5
Hydrology Lecture Note

5.3 Dimensionless Unit Hydrograph

The Soil Conservation Service (SCS) dimensionless unit hydrograph procedure is
one of the most well-known methods for deriving synthetic unit hydrographs in use
today (Note: the agency is now known as the Natural Resources Conservation
Service or NRCS, but the acronym SCS is still used in association with its UHG
method). References for this method can be found in most hydrology textbooks or
handbooks. The primary reference for this method may be considered as the Soil
Conservation Service - National Engineering Handbook, Section 4, Hydrology (SCS
1972). There are a number of versions of this reference occurring both before and
after the given date. The dimensionless unit hydrograph used by the SCS was
developed by Victor Mockus and was derived based on a large number of unit
hydrographs from basins which varied in characteristics such as size and
geographic location. The unit hydrographs were averaged and the final product was
made dimensionless by considering the ratios of q/q p (flow/peak flow) on the
ordinate axis and t/tp (time/time to peak) on the abscissa, where the units of q and q p
are flow/inch of runoff/unit area. This final, dimensionless unit hydrograph has a
time-to-peak located at approximately 20% of its time base and an inflection point at
1.7 times the time-to-peak.

Hydraulic & Water Resources Engineering Department, AMIT, AMU

6
Hydrology Lecture Note

Figure 5.3: SCS Dimensionless unit hydrograph and mass curve

For discussion purposes, this curvilinear unit hydrograph may be represented by an

equivalent triangular unit hydrograph as in Figure 5.4.

Hydraulic & Water Resources Engineering Department, AMIT, AMU

7
Hydrology Lecture Note

Figure 5.4: Illustration of dimensionless curvilinear unit hydrograph and equivalent

triangular hydrograph.

Recall that the unit hydrograph is the result of 1-inch of excess rainfall (of duration
D) spread uniformly over the basin. Using the geometry of the triangles (area = 1/2
base times height), one can calculate that the unit hydrograph has 37.5% (or 3/8) of
its volume on the rising side and the remaining 62.5% (or 5/8) of the volume on the
recession side.

Hydraulic & Water Resources Engineering Department, AMIT, AMU

8
Hydrology Lecture Note

Figure 5.5: Illustration of volume breakdown of SCS triangular unit hydrograph.

Using the dimensionless timing values on the x-axis, one can solve for the time base
in terms of the time-to-peak. The following relationships are made and will be useful
in further developing the peak rate relationships.

(5.5)

and
(5.6)

Again using the geometric relationships (area = 1/2 base times height) of the
triangular unit hydrograph, the total volume under the hydrograph is found by:

(5.7)

Hydraulic & Water Resources Engineering Department, AMIT, AMU

9
Hydrology Lecture Note

The volume, Q, is in inches and the time, T, is in hours. The peak rate, q p, in inches
per hour, is found to be :

(5.8)

The term, qp, in the above equation is converted to cubic feet per second and the
drainage
area, A (mi2), is brought into the equation, which results in :

(5.9)

f (the area under the unit hydrograph) from 1-square mile in 1-hour (3600 seconds).
Substituting in relationships developed in equation 1 above, equation 5 is rewritten :

(5.10)

Because the above relationships were developed based on the volumetric

constraints of the triangular unit hydrograph, the equations and conversions are also
valid for the curvilinear unit hydrograph, which, proportionally, has the same volumes
as the triangular representation. The conversion constant (or peaking factor) 484 is
the result of the large number of unit hydrographs from a wide range of basin
characteristics and actually reflects the ability of the watershed to retain and delay
the flow.

This constant may not be applicable to all watershed types. Steep terrain and urban
areas may tend to produce higher early peaks and thus values of the peaking factor
may tend towards 600. Likewise, flat swampy regions tend to retain and store the
water, causing a delayed, lower peak. In these circumstances values may tend
towards 300 or lower (SCS 1972; Wanielista, et al. 1997). It would be very important
to document any reasons for changing the constant from 484, effectively changing
the shape of the unit hydrograph.

When changing the shape of the unit hydrograph, one must keep in mind the ratios
of the volumes under the rising and falling sides of the original dimensionless unit
hydrograph and the resulting volume under the unit graph must remain at 1 inch.
The table below illustrates possible values for the peaking factor and the associated
ratios of recession limb length to rising limb. The data below does not include a
reference value for swampy regions; however, this would be a rather low value, as
well. Again, it is important to note that experience and judgment are very important
in determining the most appropriate values for a region.

Hydraulic & Water Resources Engineering Department, AMIT, AMU

10
Hydrology Lecture Note

Hydrograph peaking factors and recession limb ratios (Wanielista, et al. 1997)

The peak rate may also be expressed in terms of other timing parameters besides
the time-to-peak. From the previous figures:

(5.11)

where D = the duration of the unit excess rainfall and T lag = the basin lag time, which
is defined as the time between the center of mass of excess rainfall and the time to
peak of the unit hydrograph. The peak flow is now written as:

(5.12)

The SCS (1972) relates the lag time, T lag, to the time of concentration, Tc by:

(5.13)

Time of concentration is the time from the end of excess precipitation to the
inflection point on the recession limb of the hydrograph. Combining this with other
relationships, as
illustrated in the triangular unit hydrograph, the following relationships develop:

(5.14)

and

(5.15)

From this, the duration D may be expressed as:

(5.16)

Hydraulic & Water Resources Engineering Department, AMIT, AMU

11
Hydrology Lecture Note

Eq. (5.16) provides a desirable relationship between duration and time of

concentration, which will provide enough points to accurately represent the unit
hydrograph, particularly the rising limb. Eq. (5.16) IS NOT an equation for calculating
the duration of the unit hydrograph. The duration of the SCS unit hydrograph is
simply chosen as D and locks in the peak time by Eq. (5.11), above.

The peak flow rate could now be expressed in terms of the time of concentration as :

(5.17)

As a final note, for a unit hydrograph, the volume of runoff, Q, would be equal to 1-
inch.

Application of SCS Unit Hydrograph

It is necessary to estimate the lag time for a given basin (of area, A) in order to
derive the
SCS dimensionless unit hydrograph. Eq. (5.11) on the previous page:

(5.18)

The timing parameter is somewhat difficult to estimate and rather subjective;

however, this parameter has considerable influence on the values of the unit
hydrograph.

Timing Parameter(s):
There are several methods that may be considered for estimating the timing of the
unit hydrograph.

SCS Lag Approach

The SCS lag method is an empirical approach developed by the SCS for estimating
the lag time or the distance from the center of mass of excess rainfall to the peak
discharge. The SCS also recommends that the lag equation be used on basins that
are considered to be somewhat homogeneous in nature and under than 2000 acres
in size, although this may be debatable. Due to these potentially restrictive
recommendations, the method may not be applicable to NWSRFS-sized basins;
however, due to its relative ease of use and the potential to eventually have a
modeling capability for smaller basins, this method will be discussed here. The SCS
lag equation is given as:

(5.19)

Hydraulic & Water Resources Engineering Department, AMIT, AMU

12
Hydrology Lecture Note

where:
Tlag = lag time in hours
L = Length of the longest drainage path in feet
S = (1000/CN) - 10 (CN=curve number)
%Slope = average watershed slope in %
The parameter S represents the potential maximum moisture retention of the soil
and is related to soil and cover conditions of the watershed. It is empirically-
determined using an SCS curve number (CN) which can be obtained from tables in
SCS publications. The remaining parameters are the Length, L, and the % Slope.
The length, L is the length of the longest drainage path from the watershed outlet to
the watershed divide, which is generally obvious for most watersheds. The more
difficult parameter is the average slope of the watershed. This is a rather subjective
parameter, as it is highly dependent on the viewpoint of the user. It is generally
insufficient to use the slope of the longest drainage path as this will usually produce
a more mild slope and is not representative of the near-stream areas which may
have considerably higher slopes and cause faster responses.

The impact of a low value for slope is to produce a longer lag time, as can be seen
from equation 14. The longer lag time has the effect of reducing the peak of the unit
hydrograph as can be seen from equations Eq. (5.11) and Eq. (5.12).

Lindsey, Kohler, and Paulhus (1949) provide a means of estimating the average
watershed slope that involves overlaying a grid on the basin, which has been
sketched on a topographic map. The average slope is found by:

(5.20)

where:
N = total number of contour crossings on the grid lines
∆Z = the contour interval
l = the length of all the grid lines
Where N = total number of contour crossings on the grid lines, DZ = the contour
interval and l = the length of all the grid lines. In a geographic information system
(GIS), there is a grid already associated with the digital elevation model (DEM), thus
thus the average slope could be calculated by summing the number of potential
contour crossings on each row and column. The contour interval should be the
expected accuracy of the DEM (i.e 5m in the vertical). There is the potential problem
of the contour lines not being perpendicular to the rows and columns. Horton (1926)
suggested the average slope could be calculated by :

(5.21)

where q = the angle between the contours and the grid lines (rows and columns).
Due to

Hydraulic & Water Resources Engineering Department, AMIT, AMU

13
Hydrology Lecture Note

the difficulty in evaluating q, Horton suggested that an average value of sec q =

1.571
could be used. Thus equation 15 could be written:

(5.22)

In recent years, GIS have become somewhat commonplace and provide an

excellent means for calculating the watershed slopes based on either a DEM or a
triangular irregular network (TIN), or on stream segments themselves.

Segmental Approach
In the segmental velocity or segmental approach, the parameter being estimated is
essentially the time of concentration or longest travel time within the basin. In
general, the longest travel time corresponds to the longest drainage path; however,
there may be situations and basin configurations that allow for some shorter travel
distances to have longer travel times, due to roughness and/or flow type. The flow
path is broken into segments with the flow in each segment being represented by
some type of flow regime. The most common flow representations are overland,
sheet, rill and gully, and channel flow. A number of equations or graphs are available
to estimate the travel velocity of surface flow in each type of condition. The travel
times are then estimated by dividing the segment length by the associated velocity.
The sum of all of the travel times represents the time of concentration.

A common SCS relationship between time of concentration and lag time is given by:

(5.23)

The velocities in each segment may be calculated by a number of means. For

channel
segments, the velocity could be estimated via the Manning equation :

(5.24)

where:
V = velocity
n = Manning roughness coefficient
Rh = hydraulic radius
S = slope in ft/ft
This would require knowledge of the channel parameters and/or shapes for the
basin in question. An equation similar in form to the Manning equation may be
recommended :

(5.25)

Hydraulic & Water Resources Engineering Department, AMIT, AMU

14
Hydrology Lecture Note

Where k is a coefficient based on the flow type. McCuen (1989) and SCS (1972)
provide values of k for several flow situations. The values of K are assuming that
slope is given in
percent.

Coefficients of velocity (fps) versus slope (%) relationship for estimating travel
velocities (McCuen 1989; SCS 1972).

The Michigan Department of Natural Resources (Sorrell and Hamilton 1991) - Land
and Water Management Division (1991) provide very similar relationships as
illustrated below.

Coefficients of velocity (fps) versus slope (%) relationship for estimating travel
velocities (Sorrell and Hamilton 1991).

Timing Wrap-up
It is essential to have or estimate the "timing" of the unit hydrograph. The timing is
what locks in the time of the peak and helps to define the peak flow. Recall that the
lag time is from the center of mass of excess precipitation to the peak of the unit
hydrograph. Therefore; if we have a duration of 2 hours and we calculate the lag

Hydraulic & Water Resources Engineering Department, AMIT, AMU

15
Hydrology Lecture Note

time of the basin to be 6 hours, then the time to peak would be 6 hours + 1 hour (half
the duration) = 7 hours. As you can see from the equations (particularly 5.11 & 5.12),
the lag time is the desired timing, although we can also use the time of concentration
- which can be related to lag time. We have discussed only a small number of
methods of looking at the timing of a basin. There are many other equations,
although the thought process is generally the same - you must find the lag time of
the basin (or time of concentration) in order to compute the peak flow of Eq. 5.12.

Conclusions
The following items, in no particular order of importance, must be considered when
deriving the SCS unit hydrograph:
1. Slope - The slope is the average slope of the basin. This is somewhat of a
concern as the response time may not be well represented by the average slope, or
the contributing area(s) may be smaller than the overall basin area and may be
associated with a higher value of the average slope.
2. The constant 484 in the equation for peak flow of the unit hydrograph - As
mentioned this may not be applicable for all areas of the country. This will require
input from an experienced hydrologist to determine the most appropriate values for
various basins. A default value of 484 will be used.
3. Duration - The duration of the unit hydrograph plays a role in determining
the peak flow via the time to peak. Equation 5.11 provides guidance for determining
a value of the duration, and hence the computational time step, such that there are
sufficient points on the rising side of the unit hydrograph to adequately represent and
simulate runoff from a precipitation event.
4. Triangular Shape - A triangular unit hydrograph is produced. The triangular
unit hydrograph represents 1-inch of excess precipitation (the volume under the unit
hydrograph). While the curvilinear unit hydrograph is more traditional in shape; there
will be minimal difference in the overall simulation capabilities between the two unit
hydrographs. This is particularly true for single event simulations.

Overview of Computations
Below is a description of the required user inputs for the SCS method.

This could now be altered in the calibration process. Values at hourly intervals may
be
extracted or interpolated.

Hydraulic & Water Resources Engineering Department, AMIT, AMU

16
Hydrology Lecture Note

5.4 Extrapolation of flow data to Un-gauged Sites

All too often the stream flow data that are available from measured gauging stations are
not from location for which a project site analysis is to be made. Methods are required to
develop extrapolation of measured flow data which will be representative of a given site
on a stream. In regions where stream flow does not vary with respect to the contributing
drainage area flow duration curves can be plotted for the gauged sites. From these
developed flow duration curves, a family of parametric flow duration curves can be
developed, in which flow is plotted against the average annual runoff ( R ) or annual
discharge, Q at the respective gages for several exceedence interval percentages. A
separate curve is developed for each exceedence interval used. A correlation analysis is
then performed to obtain the best-fitting curve for the data taken from the measured
records of stream flow.

Figure 5.6 FDC for gauging stations in a homogenous drainage basin

Hydraulic & Water Resources Engineering Department, AMIT, AMU

17
Hydrology Lecture Note

Figure 5.7 Parametric flow duration curves

Determination of average annual discharge:

To use the parametric flow duration curves effectively, it is necessary to determine the
average annual discharge,Q, at the point or location on the stream for which a flow
analysis is to be made. Isohytal maps developed for normal annual precipitation in a
river basin are helpful for determining the annual discharge. The records of precipitation
and stream flow data should represent the same period of record.

Utilizing the records of average annual precipitation input to the basins at measured
streams nearby or having similar hydrologic characteristics, a runoff coefficient is
estimated for the drainage basin being studied. The product of this coefficient and the
computed normal annual precipitation input to the basin and the basin area can be used
to calculate the average annual discharge as:

With the average runoff annual discharge estimate it is possible to enter the parametric
flow duration curve and determine values of flow for different exceedence percentages
for which the parametric flow duration curve has been developed.

Hydraulic & Water Resources Engineering Department, AMIT, AMU

18