Flood Estimation by Unit Hydrograph Techniques

FLOOD ESTIMATION BY
UNIT HYDROGRAPH TECHNIQUES

Dr. Anil Kumar Lohani

Scientist G
National Institute of Hydrology
Roorkee-247667
Email: [email protected]
[email protected]

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

Contents
OBJECTIVES ............................................................................................................................ 2
INTRODUCTION ..................................................................................................................... 3
THE PURPOSE OF A UNIT HYDROGRAPH ........................................................................ 4
WHAT IS A UNIT HYDROGRAPH ? ..................................................................................... 4
UNIT HYDROGRAPH THEORY AND ASSUMPTIONS ...................................................... 5
SUMMARY ............................................................................................................................... 9
FACTORS AFFECTING UNIT HYDROGRAPH SHAPE ...................................................... 9
CHARACTERISTICS OF THE UNIT HYDROGRAPH AND THE DISTRIBUTION
GRAPH .................................................................................................................................... 11
DERIVATION OF UNIT HYDROGRAPH............................................................................ 11
CHANGE OF UNIT VOLUME IN A UNIT HYDROGRAPH .............................................. 23
CHANGE OF UNIT TIME PERIOD IN A UNIT HYDROGRAPH ...................................... 23
INSTANTANEOUS UNIT HYDROGRAPH (IUH) .............................................................. 23
S-CURVE HYDROGRAPH .................................................................................................... 24
RELATIONSHIP BETWEEN THE S-CURVE AND UH...................................................... 27
RELATIONSHIP BETWEEN THE S-CURVE AND IUH .................................................... 28
RELATIONSHIP BETWEEN UH AND IUH ........................................................................ 29
CHANGE OF UNIT PERIOD OF A UNIT HYDROGRAPH................................................ 29
S-CURVE METHOD .............................................................................................................. 31
AVERAGING OF UNIT HYDROGRAPHS .......................................................................... 34
CATCHMENT SIZE LIMITATION FOR UNIT HYDROGRAPH ....................................... 37
UH APPLICATION ON SMALL CATCHMENTS FOR FLOOD ESTIMATION ............... 38
FLOOD ESTIMATION FOR LARGE CATCHMENTS ........................................................ 38
APPLICATION OF UNIT HYDROGRAPH .......................................................................... 39
BIBLIOGRAPHY .................................................................................................................... 39

OBJECTIVES

After attending to this lecture, the participants would be able to know about the following
aspects:

(i) Unit hydrograph (UH) theory, its assumption and limitations;

(ii) Various factors affecting unit hydrograph shape;
(iii) Derivation of unit hydrographs;
(iv) Instantaneous unit hydrograph (IUH) and S-curve hydrograph; and
(v) Application of the unit hydrograph.

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

INTRODUCTION

Design of structures to control river flows must consider both extremes of runoff (that is,
droughts and floods). Analyses are required to size the capacity of outlet works (spillways,
bypasses, etc.) to cater for floods. It is often necessary in hydrologic design to have details of
both the peak flows and the distribution of flow with time, in other words, the hydrograph, so
that the runoff volume can be estimated.

As stream flow records are somewhat limited for most locations, it is necessary to relate runoff to
rainfall. Knowing rainfall rates, a function to convert rainfall to runoff is required. Unit
hydrograph (UH) is such a tool.

Unit hydrographs can be derived from analysis of rainfall and runoff records in those catchments
where such data are available. The procedures used to derive a unit hydrograph are dependent
upon whether the storm from which a unit hydrograph is to be calculated is:

• a simple or single period storm, or

• a multi period storm

Thunder-storms are usually intense and of short duration, and more likely to be treated as single
period storms. Frontal storms are usually of longer duration and therefore, are generally suitable
for single period analysis. A multiperiod approach should be followed for these.

Generally, the unit hydrographs derived from various events are not the same. The estimation of
design flood requires a representative unit hydrograph for the watershed. There are two possible
approaches for the derivation of the representative unit hydrograph for the watershed. In first
approach, the unit hydrographs derived from various events are averaged by the conventional
averaging method. However, the second approach considers the joint event, obtained from
clubbing the various events together for the analysis and provides a single representative unit
hydrograph for the watershed.
As mentioned about, the unit hydrograph is a conversion factor which converts the rainfall
(excess rainfall) to runoff (direct surface runoff). For this conversion, the duration of unit
hydrograph should be the same as the duration of each of the excess rainfall blocks. If the
duration of unit hydrograph and excess rainfall blocks differ, then the duration of unit
hydrograph should be changed to the duration of excess rainfall, and the unit hydrograph with
changed duration should be used for converting the excess rainfall to the direct surface runoff.
Two approaches based on S-curve and superimposition methods are envisaged for changing the
duration of unit hydrograph. The former approach of S-curve is more general than the later one
which is used for changing the duration of unit hydrograph only when the new duration is the
integer multiple of the original duration.

In this lecture most of the above aspects are discussed with the help of some illustrative
examples. If runoff data is not available (ungauged catchment case), technique have been

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

developed to derive unit hydrographs using purely catchment characteristics. These are termed
as 'regional unit hydrographs' or 'synthetic unit hydrographs'.

THE PURPOSE OF A UNIT HYDROGRAPH

Basically, the unit hydrograph is a multiplier, converting rainfall to runoff. A unit hydrograph is
a hydrograph and thus its multiplying effect varies with time, producing from rainfall a time
distribution of surface runoff. A unit hydrograph only measures direct surface runoff.
Therefore, the base flow must be separated from the stream flow hydrograph in the derivation of
a unit hydrograph and must be added to surface runoff derived using unit hydrograph techniques
to obtain total runoff.

When variations in rainfall with time, the unit hydrograph can be applied to each rainfall unit and
from an addition of the resultant individual runoff hydrographs, the surface runoff hydrographs
can be determined.

WHAT IS A UNIT HYDROGRAPH ?

(a) A unit hydrograph is a hydrograph of surface runoff that would result at a given point in
a stream from unit rainfall excess occurring in unit time uniformly over the catchment
area above that given point.
It deals only with rainfall excess and thus loss rates must be deducted from total rainfall.
(b) Earlier the normal practice was to take unit rainfall excess as one inch, with metrication,
one millimetre is now commonly used, but it may be stressed that the unit rainfall excess
can be any selected value (such as 5, 10, 20, 50, 100 mm, etc.) depending on the size of
the catchment and the rainfall magnitude.
(c) The unit time is arbitrarily selected taking into account the duration of the storm and the
size of the catchment area being examined. For small catchments use short periods, say
1, 2 hours and for larger catchments 3, 4, 6 or even 12 hours.
(d) Basically it can therefore be said:

(i) A unit hydrograph is a flow hydrograph;

(ii) A unit hydrograph is a hydrograph of surface runoff not total runoff;
(iii)The hydrograph of surface runoff results from the rainfall excess;
(iv) The rainfall excess represents total rainfall minus losses (abstractions);
(v) For any unit time period, the rainfall excess is assumed to occur uniformly over the
catchment;
(vi) Typical unit times used in unit hydrograph analysis are 1,2,3,6,8, and 12 hour
periods. Time periods which are not multiples of 24 hours, should not be used.

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

UNIT HYDROGRAPH THEORY AND ASSUMPTIONS

(a) Constant base length proposition:

The duration of surface runoff (tp) for rainfall of a know duration (tr) for one catchment is
constant and independent of the total volume of runoff (Fig. 1).

If tp = duration of excess rainfall, then for all rainfall events of this duration, the base
length of the surface runoff is constant and equal to tr.

(b) Proportional ordinate proposition:

For storms of equal duration (unit perids) on one catchment, at time t, the rate of surface
runoff is proportional to the total volume of rainfall excess.

For form 1 - rainfall excess = PE1 over time period tp = Pe1 x tp

For form 2 - rainfall excess = PE2 over time period tp = Pe2 x tp
At Time t:

q 1 (t) P e1 x t p P E1
= =
q 2 (t) P e2 x t p P E2

Example 1:

Fig. 2 shows the surface runoff hydrographs for three different excess rainfall volumes of
100 mm, 50 mm, 25 mm of same unit period. The ratios of q100:q50:q25 = 100:50:25 =

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

4:2:1.

Therefore, in unit hydrograph theory if you have a unit hydrograph derived for 100 mm
of excess rainfall and the actual excess rainfall is 60 mm in the same unit time period, the
runoff resulting is 0.6 times the unit hydrograph ordinates. If the actual rainfall rainfall
excess is 220 mm, then the runoff is 2.2 times the unit hydrograph and so on.

(c) Concurrent flow proposition:

The hydrograph of surface runoff resulting from a particular portion of storm rainfall is
not affected by concurrent runoff resulting from other portions of the storm. In other
words, the total hydrograph of surface runoff is the sum of the surface runoff produced
by the individual portions of the excess rainfall. See Fig. 3, where two hydrographs of
surface runoff are shown one resulting from excess rainfall PE1 and other from excess
rainfall PE2.

Under this principle, (concurrent flow proposition), the surface runoff (q1) from the first
rainfall excess (PE1) is not affected by the surface runoff (q2) from the second rainfall
excess (PE2). The total surface runoff is therefore the sum of q1 + q2.

In unit hydrograph derivation, it is necessary to derive a unit hydroraph which will

produce q when multiplied by PE1 and q2 when multiplied by PE2.

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

Example 2:

Fig. 4 shows the example of concurrent flow proposition where A is a given hydrograph
for a depth of net rainfall of 5 cm and a duration ªt. The hydrograph for a rain consisting
of two successive periods, each with the same duration ªt as the given ont but with depth
of excess rainfall of 6 and 3 cm respectively can be found by adding the ordinates of the
hydrographs B and C. Here B and C are the hydrographs due to the excess rainfall of 6
and 3 cms respectively.

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

(d) Uniform rainfall excess in time:

The assumption requires that:

(i) The storm selected should be of short duration and intense.
(ii) The resulting hydrograph should be a single peaked and short time base.

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

(e) Uniform rainfall excess in space:

(i) The catchment considered should be small

(ii) The unit hydrograph theory becomes inapplicable for the large watershed (more
than 5000 sq. km.)
(iii) The unit hydrograph theory can be applied by sub-dividing the large watersheds.

(f) Time invariance propositions:

This principle is valid only when:

(i) The physical characteristics of the watershed do not change will time due to man
made adjustments and land use effects.
(ii) Storm pattern and its movement do not change with time.

SUMMARY

(i) The principle of proportionality and superposition along with the assumption of uniform
rainfall in time and space constitute the unit hydrograph theory.
(ii) The principle of proportionality states that for the same unit period of excess rainfall the
surface runoff hydrograph is directly proportional to the excess rainfall volume.
(iii) When correctly sequenced with respect to the timing of therainfall, surface runoff
hydrograh can be superimposed and added to arrive at total surface runoff. The term
used to correctly sequence the surface runoff hydrograph is lagging.
(iv) An isolated, intense, short duration storm or nearly uniform distribution in space and time
is the most desirable for the unit hydrograph derivation. However, such storms are rare
in practice. Therefore, one has to use the complex storms for deriving the unit
hydrograph.
(v) The shapes of the unit hydrograph are strongly dependent on the effective storm pattern
and the physical condition of the watershed prior to the storm. As a result unit
hydrographs derived from different storms of the same watershed may vary.

FACTORS AFFECTING UNIT HYDROGRAPH SHAPE

The two basic factors that affect the shape of the unit hydrograph are:

(a) Rainfall distribution over the catchment

(b) The physiography of the catchment - its shape, slope, vegetation cover, soil type, etc.

(a) Rainfall distribution:

The primary variations in unit hydrographs, derived from different storms, are due to
variations in areal pattern of rainfall, rainfall duration and time intensity pattern.

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

A hydrograph resulting from precipitation concentrated in the lower part of a basin will
have a rapid rise, a sharp peak and a rapid recession. Precipitation concentrated in the
upper part of the same basin will have a slower rise and recession and a lower, broader
peak. The unit hydrographs developed from different areal distributions of runoff would
have very different shapes. The differences in spatial distribution of rainfall may be
considerably reduced if the application of the unit hydrograph technique is restricted to
small basins.

The limitation of small basins does not ordinarily apply rainfall variations caused by
topographic effects, since the effects can be considered as relatively fixed characteristics
of the basin.

Various from the normal areal pattern of precipitation cause the differences in unit
hydrographs. The time base of the unit hydrograph will be lengthened and the peak will
be lowered as the duration of rainfall is increased for the same amount of runoff. A
separate unit hydrograph is theoretically necessary for each possible duration of rainfall.
Time intensity pattern of rainfall can have a significant effect on unit hydrographs and
the effect is directly related to basin size on large basins, where changes in storm
intensity must last for several hours to cause distinguishable effects on the hydrograph.
The short bursts of rainfall lasting only for few minutes may cause the clearly defined
peaks in the hydrographs of very small basins.

For large basins, valley storage tends to eliminate the effects of short time intensities and
only major changes in the time intensity pattern are reflected in the hydrograph. The
effects of changes in the time intensity pattern can usually be lessened by selecting the
computational interval to be used in developing a unit hydrograph short enough so that
the changes are not large from one computational interval to the next.

(b) Physiography of the catchment:

Changes in the physical characteristic do occur from natural and man made causes. As a
result, the shape of the unit hydrographs also changes.

Catchments of the same size but with different physiographic factors produce differently
shapes hydrographs. Steep catchment slopes result in faster runoff and thus tend to have
more peaked unit hydrographs with the peaks occuring earlier than catchments with
flatter slopes.

Urbanization of a catchment causes drastic changes in the shape of hydrographs and unit
hydrograph developed. Due to urbanisation, the natural valley storage of the basin and
the average loss rates would be reduced. As a result, the unit hydrograph will tend to
have higher peaks and shorter times of concentration.

Under natural conditions, changes in physical characteristics can occur due to seasonal

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

and long term changes in vegetation or to other causes, such a fires, etc.

Some regional relationship, between the unit hydrograph parameters and existing basin
characteristics, can be utilised to derive the unit hydrographs for the basin due to change
in the catchment characteristics.

CHARACTERISTICS OF THE UNIT HYDROGRAPH AND THE DISTRIBUTION

GRAPH
The main characteristics of the unit hydrograph and the distribution graph are:

(i) It gives the time distribution of the discharge hydrograph of a watershed produced by a
uniform net rain of given depth precipitated on the area.
(ii) It shows how this rain is transformed into discharge at the outlet. This transformation is
assumed to be a linear process.
(iii) It is a characteristic for a given watershed, it shows the integrated effect of the surface
features on the routing of the rain through the catchment.
(iv) To derive a unit hydrograph, it is necessary to dispose of a number of records of, at least,
moderate floods.
(v) The principles of the unit hydrograph can be applied for estimating the design flood,
supplementing missing flood records and short term flood forecasting based on recorded
rainfall.
(vi) The distribution graph shows the percentage of total unit hydrograph occurring during
successive time periods. The sum of all the percentage must be equal to 100.

DERIVATION OF UNIT HYDROGRAPH

Conventional Method
The unit hydrograph for a gauging station is derived as flows from analysis of recorded floods
for the single period storm:
(i) Derive the total hydrographs for the major floods from stream flow records using gauge
height flow ratings (rating curves);
(ii) The base flow should be separated from the total hydrograph to produce the direct
surface runoff hydrograph and thus the surface runoff volume Q in millimetres. In
practise it is usual to analyse floods which produce a result for Q, greater than 20 mm.
Often this is not possible and lower values can be used, but with some caution.
(iii) Rainfall records for the stations in the catchment must be analysed and the average
rainfall for the storm which produced that flood determined using one of the procedures;
(iv) The hydrograph of the total average rainfall is derived for time periods equivalent to the
required unit period of the unit hydrograph;
(v) Loss rates are evaluated to apply to the total rainfall to estimate the excess rainfall in
each of the unit periods.

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

(vi) Having determined the hydrographs of excess rainfall and confirming that the excess
rainfall occurs in only one single unit period, the derivation of the unit hydrograph is
quite simple.
(vii) Remembering that the unit hydrograph is a multiplying function to convert excess
rainfall to surface runoff, the following procedure is followed:
• Plot the surface runoff hydrograph
• Determine the volume of excess rainfall, PE , in the single unit period. (PE also
equals the volume of the surface runoff hydrograph Qs);
• Relate the value of PE to the unit volume required in the unit hydrograph, QUH

For example, if PE = 22.5 mm and you want to derive a unit hydrograph for 10 mm (QUH = 10
mm), then the proportionality factor (F) used to convert the surface runoff hydrograph ordinates
to unit hydrograph ordinates are:

22.5
F = PE = = 2.25
QUH 10

22.5
F= = 0.450
50

If the unit hydrograph is to represent 50 mm then:

Divide the surface runoff hydrographs by F and this gives the unit hydrograph ordinates.

Example 3:
The thunder storm of six hours durations with the excess rainfall of 154 mm produces the
following surface runoff hydrograph:

Date & Time (Hrs) Surface Runoff (m/s)

June 15, 0600 10
1200 500
1800 1600
M.N. 3500
June 16, 0600 5200
1200 3100
1800 1500
M.N. 650
June 17, 0600 250
1200 0
1800 0
M.N. 0

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

Find out the ordinates of 6 hour unit hydrograph having unit volume equal to 1000 mm.
Solution:

A unit hydrograph for 100 mm is required.

Hence,

PE = 154, QUH = 100

154
_ F= = 1.54
100

The required unit hydrograph ordinates are found dividing the surface runoff hydrograph
ordinates by the factor F = 1.54. The calculation of 6 hour unit hydrograph for the above storm
is shown in Table 1.

Table 1 : Computation of 6 hour unit hydrograph

Date Time Surface Runoff PE 6 hour 100 mm unit

(Hrs) (m3/s) (mm) hydrograph (m3/s)
(Q) (Q/F)
June 15 0600 10 154 7
1200 500 325
1800 1600 1039
2400 3500 2272
0600 5200 3377
1200 3100 2013
1800 1500 974
2400 650 422
0600 250 162
1200 0 0
1800 0 0

Derivation of unit hydrographs from Multi period storms

When a storm is of short duration and of fairly uniform intensity, the unit hydrograph can be
derived simply by applying the single period storm technique.

For a storm of long duration or a shorter one with variable intensity, the storm must be treated as
consisting of a series of storms and the derivation of the unit hydrograph is slightly more
complex.

There are number of methods to derive unit hydrographs from multiperiod storms but all are

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

based on unit hydrograph theory.

Let:

Pej = Excess rainfall intensity (mm/hour)

PEi = Excess rainfall volume (mm)
Ul = Ordinates of Unit hydrograph (m3/sec)
qi = Ordinates of surface runoff hydrograph (m3/sec)

Before illustrating how to derive unit hydrographs for multi period storm, the following example
shows how the unit hydrograph theory is used to produce the surface runoff hydrograph from a
multi period storm.

If Pei = Rainfall rate in hours in period Tl (mm/hour)

then PEi = volume of rainfall = Pel x Tl
e.g. If Pe1 = 2 mm/hour
T1 = 2 hours
ˆ PE1 = 2 x 2 = 4 mm

To determine the surface hydrograph from 4 mm of rainfall in2 hours, we use the unit
hydrograph of 2 hour duration which has 7 unit period (7 x 2 = 14 hours). Then in unit
hydrograph theory:

(i) The base length of the surface runoff for rainfall of 2 hours duration is 14 hours.
(ii) For a 14 mm unit hydrograph, under the proportionality principle, the average runoff in
the various periods for the first rainfall excess of 4 mm, is:

q1 = 0.4 x U1 q5 = 0.4 x U5
q2 = 0.4 x U2 q6 = 0.4 x U6
q3 = 0.4 x U3 q7 = 0.4 x U7
q4 = 0.4 x U4

(iii) If Pe2 = 4 mm/hour (2nd period of rainfall)

PE2 = 4 x 2 = 8 mm

Then the ordinates for the second period of rainfall are:

q1 = 0.8 x U1
q1 = 0.8 x U2, etc.

However, the rainfall in the second period started (lagged) T1 (2 hours) later and thus q1 for the
second period of rainfall must be added to the second period of runoff for first hydrograph.

This principle continues for the third and subsequent rainfalls.

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

The equations for the combined surface runoff can therefore be generalised as follows (Fig. 5):

q1 = PE1 U1 Ordinates of Surface

q2 = PE1 U2 + PE2 U1 Runoff Hydrograph
q3 = PE1 U3 + PE2 U2 + PE3 U1
q4 = PE1 U4 + PE2 U3 + PE3 U2
q5 = PE1 U5 + PE2 U4 + PE3 U3
q6 = PE1 U6 + PE2 U5 + PE3 U4
q7 = PE1 U7 + PE2 U6 + PE3 U5
q8 = + PE2 U7 + PE3 U6
q9 = 0 + 0 + PE3 U7

All the procedures used for derivation of unit hydrographs from multi-period storms use the
principles of proportionality and superposition which are evidenced in the above equations.

(a) The method of single division:

This procedure is basically a trial and error approach. The procedure is to solve each of
the equations relating to the unit hydrograph ordinates (Ref. Fig. 5).

(i) For the first rainfall both q and P are known, q from the surface runoff
hydrograph and P from the hyetograph of rainfall:

If q1 = PE1 x U1
q
U1 = 1 amd thus U1 is found.
P E1
(ii) For the second rainfall, q2 is known from the hydrograph and PE1, U1 and PE2 are
known:

Since q2 = PE2 U2 + PE2 U1

U 2 = q 2 - P E2 U 1overP E1
(iii) Continue through the set of equations. However, since there are more equations
than unknown, once you have worked through you will find usually that the
equations for qs and q9 will not be solved by using U6 and U7 ordinates
determined from q6 and q7 equations.

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

It is necessary to apply an 'averaging' procedure and this involves calculating the values
of U1, U2, etc. upto U7 from a 'forward run', that is, from q1, q2, etc. in sequence to q7.
The values of U7, U8, etc. to U1 are then determined from q9, q8, etc. in sequence to q1 by
reverse calculations.
Fig. 6 illustrates typical results obtained. The first approximation of the unit hydrograph
is obtained by 'average' the values from the two runs.

Test the unit hydrograph by checking how the derived hydrograph compared with the

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

recorded hydrograph - THIS TEST SHOULD BE DONE IN ALL UNIT

HYDROGRAPH DERIVATIONS.

When testing the unit hydrograph obtained by averaging, it is some times found that the
reproduction is not good. This can be due to many things such as a small error in U1 and
in U2 and in U6 and U7, which are the leading and trailing ordinates, can magnify errors
in later calculations. For averaging place more reliance on U1 and U2 from forward runs
and U6 and U7 for reverse runs. The records of rainfall and runoff are inadequate and do
not represent the actual situations to any acceptable degree. Unit hydrograph theory
itself may not apply to the storm. Storms which are concentrated rather than uniformly
distributed over the catchment tend to make unit hydrograph derivation in this way
difficult.

(b) The Least Squares Method: As evidenced in the single Division Method, there are
more equations than unknown and the solution adopted depends on which set of
equations are used.

The least squares method is a statistical curve fitting procedure which provides a 'line of
best fit'. The equations which provide the estimates of unit hydrograph ordinates using
this method are described in Appendix-1 in matrix notations. The basic principle of the
least square method is the fact that the line of best fit produce the best reproduction of the
hydrograph when the unit hydrograph is re-applied to the rainfall excess. (Ref Fig. 7).
The unit hydrograph adopted is the one that satisfied the statistical objective of
minimising the sum of the squares of the deviations of the ordinates of the reproduced
hydrograph from the ordinates of the actual hydrograph.

As the method involves some tedious computations to satisfy the statistical objective, it is
not a practical proposition to carry out least squares derivations by hand. Computers are
used for this, and numerous package programmes are readily available. It is probably
one of the better methods, provided that:

• The hydrograph is representative of the basin;

• The rainfall and hydrograph ordinates are accurate;
• The unit hydrograph theory applied to the catchment.

If these criteria are not met, the shape of the 'best fit' unit hydrograph can be strange and
ordinates can even be negative. It must be stressed that the Unit Hydrograph so derived
is the mathematical multiplying function and does not necessarily represent the physical
constraints of runoff peculiar to the particular catchment. Kutchment (1967) stated a
procedure to derive smooth unit hydrograph. It involves a regularization factor, %,
which is to be determined by trial and error procedure.

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

(b) Collin's method:

This is also a trial and error solution. Assuming that the unit hydrograph has four units
of time T2 (U1, U2, U3, and U4), and there are three periods of time T of excess rainfall
(PE1, PE2 and PE3) with PE2 the maximum rainfall volume, then the surface runoff
ordinates are determined as described earlier in section 3.2 of this lecture. In the
previous example the unit hydrograph had 7 ordinates whereas in this example there are
only 4. It is not necessary when converting rainfall to runoff as shown in the example for
the selected unit hydrograph ordinates to be over the same time interval as the rainfall
excess. All that is important is that the unit hydrograph represents runoff for the excess
rainfall of the same interval as that of the excess rainfall which is being examined.

q1 = PE1 U1
q2 = PE1 U2 + PE2 U1
Ordinates of q3 = PE1 U3 + PE2 U2 + PE3 U1
Surface Runoff q4 = PE1 U4 + PE2 U3 + PE3 U2
Hydrograph q5 = PE1 U5 + PE2 U4 + PE3 U3
q6 = PE1 U6 + PE2 U5 + PE3 U4

The surface runoff hydrograph is shown in Fig. 8.

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

Procedure:

The following steps may be adopted for the unit hydrograph derivation by Collin's
method for a gauged catchment:

(i) Derive the total hydrographs for the major floods from stream flow records using
rating curves.
(ii) Estimate the surface runoff separating the base flow using one of the procedures
described in earlier lecture No 5.
(iii) Estimate the average rainfall hyetograph using the procedure described in earlier
lecture No 3.
(iv) Compute the effective rainfall hydrograph separating the losses from total rainfall
hyetograph.
(v) Ignore the terms in the equation that contain the maximum rainfall (in this
example P, termed PEMAX).
(vi) Assume a first trial unit hydrograph, that is, a set of U values that look reasonable
(use either constant value for unit hydrograph ordinates or single division

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

procedure for first estimate of unit hydrograph). Put this set of values in column 3
of Table 2.
(vii) Apply the intial trial unit hydrograph to all periods of rainfall excess except the
maximum and determine the hydrograph that would result. Refer to Table 2 and
this illustrates what this application gives (See col. 4).
(viii) If we deduct from the actual surface runoff hydrograph the sum of P and P
runoffs, we get an estimate of the runoff from maximum rainfall PEMAX. See
col. 6 in Table 2.
(ix) Dividing these ordinates by PEMAX, we get another estimate of the unit
hydrograph ordinates. See col. 7.
(x) Compare these unit hydrograph ordinates with the original trial ordinates U for
PEMAX occurs at the same time.
(xi) If the comparison is not reasonable, average the unit hydrograph ordinates of the
first trial and those derived to give second trial values as follows:
M Ui + N Ui
Ui=
M + N
where,
M = total rainfall excess except the largest one
N = largest block of rainfall excess.
It is important with trial ordinates of the unit hydrograph to ensure the unit runoff
volume corresponds to that requires (this can be any multiple of 1 mm).
(xii) Repeat step (vii) to (xi) until the calculated unit hydrograph agrees with the trial
unit hydrograph.
In Table 2 the various terms used are described as follows:
Ui = Average trial unit hydrograph ordinate in period i
qi = Average surface runoff ordinate in period i
PEi = Excess rainfall volume in period i
Ui = Adjusted trial unit hydrograph ordinate in period i
Ui = (M Ui + N Ui)/(M + N)
M = PE1 + PE3, N = PEMAX

FE = ∑ U i , where UUH = Unit hydrograph unit volume

U UH
and 3Ui = sum of solumn (8) between limits of PEMAX influence.

Example 4:

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

The excess rainfall and surface runoff ordinates for a storm of a typical catchment
are given below:

Time Periods P q average

(6 hrs) (mm) (m/s/6 hrs)
1 40 250
2 100 1050
3 60 2050
4 4350
5 4150
6 2300
7 1070
8 450
9 120

The area of the catchment is 170,000 hectares. Find out the 6 hours unit
hydrograph using Colling's method.

Solution:

Table 3 and 4 presents two trial runs on Collin's method. The objective is to find
a 6 hour unit hydrograph for 100 mm of surface runoff.

Briefly the table consists of:

(i) Indication of time period either in hours or number of periods.

(ii) Rainfall excess determined from anlaysis of recorded rainfall and runoff.
(iii) Trial unit hydrograph ordinates- The number of periods in the unit
hydrograph is determined from a combination of the number of periods of
excess rainfall and the number of periods of surface runoff (Principl of
superposition). For example, if there are 7 units on the surface runoff
hydrograph and only 1 period of rainfall, the unit hydrograph has 7 units.
If 2 periods of rainfall produce the surface runoff, then there are 6 unit
hydrographs.

The general equation for the base length relationships is:

periods in surface runoff base = Periods in unit hydrograph base +

number of rainfall periods - 1
or conversely:
Periods in unit hydrograph base = Periods in surface runoff base - number
of periods + 1

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

e.g. (1) U.G. base = 7 - 1 + 1 = 7 (for 1 period of rainfall)

(2) U.G. base = 7 - 2 + 1 = 6 (for 2 periods of rainfall)

In Table 3:

Surface runoff base = 9 (six - hour periods)

Rainfall periods = 3 (six - hour periods)
ˆ U.G. base = 9 - 3 + 1 = 7 six hour periods (42 hrs)

In table 3, the first trial assumes a uniform unit hydrograph and this is
determined as follows:

Calculate total surface runoff (from volume under hydrograph) = 340 x

106 m3
Catchment area = 170,000 hectares
= 1700 x 10 sq. metres.

For unit runoff of 1 mm unit hydrograph volume = 1.7 x 106 m3

For unit runoff of 100 mm unit hydrograph volume = 170 x 106 m3

340 x 106
Excess runoff in hydrograph = = 200mm
1.7 x 106

PE1 = 40 mm , P E1 = 0.40
100

PE2 = 100 mm , P E2 = 1.0

100

PE3 = 60 mm , P E3 = 0.6
100
Total = 200 mm.

First trial assume constant flow in unit hydrograph.

First trial average:

170 x 106
Ui= = 4.05 x 106 m3 /hour
42
6
10
= 4.05 x = 1125 m3 / sec
3600

(iv) Repeat the stated procedure unit %U (adjusted) are almost equivalent to

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

trial units than adopt as unit hydrograph.

CHANGE OF UNIT VOLUME IN A UNIT HYDROGRAPH

To convert unit hydrographs from X mm (say QUG = 100 mm) to Y mm ( say QUG = 50 mm)
X 100
i.e. = =2
Y 50

CHANGE OF UNIT TIME PERIOD IN A UNIT HYDROGRAPH

Simply divide the ordinates of the unit hydrograph (X mm) by the ratio of the volumes.
The time period should be determined using the following criteria:

(i) It should be short enough to define the hyetograph and the hydrograph with reasonable
precision.
(ii) It should be a simple fraction of 24-hours do not use 5-hours or 7 hours etc. or periods
that are not whole fraction of 24 hours.
(iii) A convention adopted in practice is to use a unit period approximately 1/3 to 1/4 the
period of rise of the hydrograph. (This is not always practical).

INSTANTANEOUS UNIT HYDROGRAPH (IUH)

If the period of a unit hydrograph is shortened but the depth is kept constant, the peak flows will
increase and the base length will decrease. If the period becomes zero and depth remains the
same, the intensity of the rain becomes infinite. This corresponds with the instantaneous
application of a sheet of water over entire catchment area. This water drains off by gravity and
the resulting unit hydrograph (IUH) Fig. 9.

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

Thus, the instantaneous unit hydrograph is a purely theoretical concept and represents the unit
hydrograph obtained when the unit excess rainfall volume occurs instantaneously. IUH
represents a more characteristics curve of a catchment area than a T-hour unit hydrograph as it is
not affected by the duration. It is a useful tool in regional unit hydrograph studies.

S-CURVE HYDROGRAPH

An S-curve is the hydrograph of direct surface runoff that would result from excess rainfall of
unit volume occurring per unit period continuously (See Fig. 10) S-curve may be derived using
the following steps:

(i) Plot the unit hydrograph of unit volume U1 at time t = 0

(ii) Plot the unit hydrograph successively lagged time T1 which equals the unit period of unit
hydrograph.
(iii) At time tb which is the base length the first unit hydrograph the S-curve becomes
constant.
(iv) The S-curve for the unit period (T1 hours) has been determined. It is generally noticed

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

that the S-curve, some times, fluctuates in the upper range. These fluctuations are
commonly referred to as a 'hunting effect', and this is characteristics of S-curve
derivation. The procedure to be followed is to plot a curve of best fit through the points
and use the values from the curve ordinates.

0.2778 AU 1
Q max =
T1

(v) The maximum ordinate of the S-curve corresponds with the equilibrium discharge:
where, A is the catchment area (sq. km.), U1 is the unit volume (mm), and T1 is the
duration of the unit hydrograph (hours).

The derivation of S-curve from a given unit hydrograph is explained in the following example:

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

Example 5:
For a catchment, ordinates of 6-hours unit hydrograph with volume 10 mm are given below:

Dated Time Interval 6-hour unit

(hours) (hours) hydrograph volume 10
mm (m/s)
June 15 0600 0 0
0900 3 200
1200 6 500
1500 9 1000
1800 12 1600
2100 15 2400
2400 18 3500
June 16 0300 21 4200
0600 24 5200
0900 27 4400
1200 30 3100
1500 33 2300
1800 36 1500
2100 39 1000
2400 42 650
June 17 0300 45 400
0600 48 250
0900 51 150
1200 54 0
If the equilibrium discharge is 16250 cumec. compute the S-curve ordinates.

Solution:
The steps are: (Ref. Table 5)

(i) Enter the ordinates of 6-hour unit hydrograph in col. 4.

(ii) Lag the 6-hour unit hydrogrph by 6-hour, 12 hour, etc. upto 54 hours, (the time base of the
unit hydrograph) and enter these values in col. (5) to (13).
(iii) Compute the S-curve addition summing the values of col. (5) to (13) at each interval.
These values are shown in col. (14).
(iv) Plot the S-curve obtained at step (iii).
(v) The plotted S-curve shows some hunting effect in the upper range. Plot a best fit curve
through these points in order to get required S-curve. Fig. 11 shows the smooth S-curve.
(vi) Read the values of the S-curve, obtained at step (v), and enter in the table as col. (15). This
is the S-curve of intensity 10/6 mm/hr.
(vii) Multiply the ordinates of col. (15) by 6 to get the S-curve of intensity 10 mm/hr. (col. 16).

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

RELATIONSHIP BETWEEN THE S-CURVE AND UH

A simple relation exists between the S-curve of a watershed and its UH. In fig. 12, St is the S-
curve for a continuous rain of i cm/hr. and St-T is the same curve shifted over a distance of T-
hours to the right. This means that the difference in the ordinates of the two (S and S ) represent
the ordinates of a hydrograph resulting due to excess rainfall depth of T cm occuring over the
catchment for Ti hours. Therefore:

U depth i T (T,t) = St - St-T

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

S t -T
or U depth d (T,t) = S - d
iT

d S t - S t -T
=
i T

here,d = depth of the T-hour unit hydrograph

i = net rainfall intensity belonging to the S-curve.
if d = 1, i = 1

S t - S t -T
ˆ U (T, t) =
T

where S is the S-curve of unit intensity (1 cm/hr.).

RELATIONSHIP BETWEEN THE S-CURVE AND IUH

S t - S t -T
Since U depth d (T, t) = d
T
As T approaches zero, U, (T,t) approaches to U (O,t) and U depth d (T, t) approaches to

d St d
dt i
d St d
ˆ U (O,t) = u (t) =
dt i

t
i
d ∫o
_ St = u(t)dt

If i = 1 and d = 1, the above equation becomes

S t = ∫ u(t)dt
.

Figure 13 shows the IUH and corresponding S-Curve.

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

RELATIONSHIP BETWEEN UH AND IUH

The relationship between the IUH (u(t), and the T-hours unit hydrograph (U (T,t), both with the
same unit depth can be found by again using the two S-curve (Fig. 14).

S t = ∫ u(t)dt
.

T.U(T,t) = St - St-T

t t
= ∫ u(t)dt - ∫ -T u (t)dt
o o

t
=
t -T
∫ u(t)dt

= A' (Fig. 14).

The T-hour unit hydrograph with a unit depth is given by:

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

A′
U (T,t) =
T

A′
Since is approximately equal to, a T-hours unit hydrograph can be easily derived from an
T
IUH. If the instantaneous unit hydrograph is available, the ordinate of the T-hour unit
hydrograph (1,2,3 hrs, etc.) at the end of the time unit T is equal to the average ordinate of the
instantaneous unit hydrograph over the T-hour period. A typical unit hydrograph of T-hour
duration, obtained in this way, is shown in Fig. 15.

Dr. A.K. Lohani, Scientist G, National Institute of Hydrology, Roorkee-247667

 Flood Estimation by Unit Hydrograph Techniques

CHANGE OF UNIT PERIOD OF A UNIT HYDROGRAPH

Having derived a unit hydrograph for a particular unit period (say 6 hrs) and you want to change
the period the following procedures can be used:

Superimposition method: - Only suitable when the new duration of unit hydrograph is integer
multiple of the original duration (say original duration 6-hrs and new duration 12 hrs).

S-curve method: - more general method.

Superimposition method

The unit hydrograph of 2 t, duration can be derived from a unit hydrograph of t, duration in the
following steps:

Add the ordinates of t - hour unit hydrograph to the ordinates of an identical unit hydrograph
lagged by t - hour.

Divide the ordinates of the resulting hydrograph of step (i) by 2 to obtain a unit hydrograph for a
unit duration of 2 t. Fig. 16 shows conversion of a unit hydrograph of t duration to duration 2 t.

Note that the unit hydrograph of n t, duration can be derived by n time successive lagging of the t
duration unit hydrograph and then dividing the resulting hydrograph by n, where n is an integer
(n = 1,2...etc.).

Workshop on Flood Management, January 18-22, 2010, National Institute of Hydrology, Roorkee-247667 L-7 29/40
 Flood Estimation by Unit Hydrograph Techniques

Example 6:
The ordinates of 6-hour unit hydrograph are given below:
Time (Hrs) 6-hour unit hydrograph ordinates (m/s)
0 0
3 200
6 500
9 1000
12 1600
15 2400
18 3500
21 4200
24 5200
27 4400
30 3100
33 2300
36 1500
39 1000
42 650
45 400
48 250
51 150
54 0

Derive 12-hour unit hydrograph of the same unit volume as of the 6-hour unit hydrograph.

Solution:

Workshop on Flood Management, January 18-22, 2010, National Institute of Hydrology, Roorkee-247667 L-7 30/40
 Flood Estimation by Unit Hydrograph Techniques

The computational steps are: (Ref. Table 6)

(i) Enter the ordinates of 6-hour unit hydrograph in col. (2).

(ii) Lag the 6-hour unit hydrograph by 6-hour and enter these values in col. (3).
(iii) Add the respective ordinates of column (2) and col. (3). The resulting hydrograph
ordinates are shown in col. (4).
(iv) Divide the ordinates of the hydrograph, obtained from step (iii), by 2 to get required unit
hydrograph of 12-hour duration. The ordinates of this unit hydrograph is given in col. (5).

Table 6: Derivation of 12-hour UH from 6-hour UH

Time (hrs) 6-hr unit 6-hr unit Superimposed 12-hour unit

hydrograph hydrograph hydrograph (m/s)
ordinates (m/s) ordinates lagged (m/s)
by 6 hour (m/s)
(1) (2) (3) (4) (5)

0 0 - 0 0
3 200 - 200 100
6 500 0 500 250
9 1000 200 1200 600
12 1600 500 2100 1050
15 2400 1000 3400 1700
18 3500 1600 5100 2550
21 4200 2400 6600 3300
24 5200 3500 8700 4350
27 4400 4200 8600 4300
30 3100 5200 8300 4150
33 2300 4400 6700 3350
36 1500 3100 4600 2300
39 1000 2300 3300 1650
42 650 1500 2150 1075
45 400 1000 1400 700
48 250 650 900 450
51 150 400 550 275
54 0 250 250 125
57 - 150 150 75
60 - 0 0 0

S-CURVE METHOD

After deriving the S-curve of unit intensity (or some other intensity), from t duration unit

Workshop on Flood Management, January 18-22, 2010, National Institute of Hydrology, Roorkee-247667 L-7 31/40
 Flood Estimation by Unit Hydrograph Techniques

hydrograph the unit hydrograph of t duration can be obtained as follows:

(i) Shift the S-curve by t hours to get the curve S (Fig. 17)
(ii) Subtract the S curve from S curve giving another curve.
(iii) The difference between the two S-curves represent the unit hydrograph for time t with a
unit volume equal to t/t of U, where U is the unit volume of the original duration unit
hydrograph.
(iv) To produce a unit hydrograph for the period t with a unit volume U multiply the
difference between the S-curves by t/t.

Example 7:

The S-curve ordinates of intensity 1 cm/hour, derived from y hour 1 cm unit hydrograph are
given below:

Time (Hrs) S-curve hydrograph ordinates (m/s)

0 0
3 1200
6 3000
9 7200
12 12600
15 21600
18 33600
21 47400
24 62400
27 74400
30 82800
33 88200

Workshop on Flood Management, January 18-22, 2010, National Institute of Hydrology, Roorkee-247667 L-7 32/40
 Flood Estimation by Unit Hydrograph Techniques

36 91800
39 94200
42 95400
45 96600
48 97200
51 97380
54 97500
57 97500

Derive a unit hydrograph of 3-hour duration with 1 cm volume.

Solution:

The computational steps are: (Ref. Table 7)

(i) Enter the S-curve ordinates in col. (2).

(ii) (S-curve is derived from 6-hour UH in the eample 5 of S-curve derivation)
(iii) Shift the S-curve by 3 hours and enter these values in col. (3).
(iv) Subtract the two S-curves [col.(2) and (3)] and enter these values in col. (4). This
represents the unit hydrograph of duration 3 hours with a unit volume equal to 3 cm.
(v) Compute the unit hydrograph of duration 3 hour with a unit volume equal to 1 cm after
dividing the hydrograph obtained from step (iii) by 3. The ordinates of the resulting unit
hydrograph is given in col. (5).

Table 7: Derivation of 3-hour Unit Hydrograph from 6-hour Unit Hydrograph S-curve
Method

Time (hrs) S-curve S-curve lagged 3-hour unit 3-hour unit

ordinates of 3-hrs hydrograph with hydrograph with
intensity 1 (m/s) volume 3 cm volume 3 cm
cm/hr (m/s) (m/s)
(m/s)
(1) (2) (3) (4)=(2)-(3) (5)=(4)/3)
0 0 - 0 0
3 1200 0 1200 400
6 3000 1200 1800 600
9 7200 3000 4200 1400
12 12600 7200 5400 1800
15 21600 12600 9000 3000
18 33600 21600 12000 4000
21 47400 33600 13800 4600
24 62400 47400 15000 5000
27 74400 62400 12000 4000
30 82800 74400 8400 2800
33 88200 82800 5400 1800

Workshop on Flood Management, January 18-22, 2010, National Institute of Hydrology, Roorkee-247667 L-7 33/40
 Flood Estimation by Unit Hydrograph Techniques

36 91800 88200 3600 1200

39 94200 91800 2400 800
42 95400 94200 1200 400
45 96600 95400 1200 400
48 97200 96600 600 200
51 97380 97200 180 60
54 97500 97300 120 40
57 97500 97500 0 0

AVERAGING OF UNIT HYDROGRAPHS

In practice, when unit; hydrographs are derived for a particular catchment, the unit hydrographs
that result will not be the same. The unit hydrographs are thus usually averaged.

Text Book Method of Averaging

The text book method (e.g. Wilson (1974), Linsley et al (1975) of averaging is performed in the
following steps:

(i) Plot the unit hydrographs, derived from a number of individual events, on a single diagram.
(ii) Mark a point which corresponds to average peak ordinates and average time to peak (Fig.
18).
(iii) Sketch an average unit hydrograph by eye so that it passes through the average peak point,
has unit volume, and generally conforms to the characteristics shape of the individual unit
hydrographs.
(vi) The procedure relies on a limited amount of trial and error (to preserve volume) and on a
good deal of subjective judgement.

Workshop on Flood Management, January 18-22, 2010, National Institute of Hydrology, Roorkee-247667 L-7 34/40
 Flood Estimation by Unit Hydrograph Techniques

Ordinate by ordinate averaging

Various ordinate by ordinate averaging methods are possible and four such schemes are:
(a) Mean method
(b) Median method
(c) Mean peaks aligned method
(d) Median peaks aligned method

(a) Mean method: In this method the average unit hydrograph ordinates are computed by taking
the mean of the corresponding ordinates of individual unit hydrographs.

Fig. 19 shows the average unit hydrograph computed by this method. It can be seen from the
Fig. 19 that the average unit hydrograph so derived tends to have a low peak and could
produce a lower hydrograph. However, as far as possible you should apply the unit
hydrograph to the excess rainfall to reproduce the surface runoff hydrographs for the
recorded storm used for unit hydrograph derivation.

The average unit hydrograph should also be reapplied to see how it reproduce the actual
hydrographs.

(b) Median method: This method computes the average unit hydrograph by ordinate
averaging. This gives a better representation of unit hydrograph shape, but if time to peak is
important, it is essential that the rising limb is adjusted so that the commencement of runoff
becomes the critical characteristics.

For unit hydrographs so that peak coincides, and then average the rising and falling limbs
check the volume to maintain unit volume. Fig. 20 illustrates this.

(c) Median peaks aligned method: In this method the steps are:

•Plot unit hydrographs so that peaks coincide and then compute the median of the rising
and falling limbs.
• Check the volume to maintain unit volume.
Shape factor averaging

Workshop on Flood Management, January 18-22, 2010, National Institute of Hydrology, Roorkee-247667 L-7 35/40
 Flood Estimation by Unit Hydrograph Techniques

The basis of the approach is the characterization of unit hydrographs by statistics namely the
volume, the mean, the coefficient of variation, skewness and peakedness. Strictly speaking, the
volume and mean characterize the scale and location of the distribution (namely the unit
hydrograph) rather than its shape, but the five statistics collectively referred as shape factors for
convenient only.

The method of using shape factors to determine an average unit hydrograph is as follows:

(i) Calculate the shape factors values for individual derived unit hydrograph and average the
values. (A median form of averaging again has the merits of minimizing the effect of
outliers).
(ii) The shape factor values of the individual unit hydrographs are then scanned until the unit
hydrograph is noted that has shape factors approximately equal to the average values.
(iii) The unit hydrograph obtained from step (ii) would represent the average unit hydrograph
for the catchment.

Joint Analysis Technique

The joint analysis technique avoids the two stage process of first deriving unit hydrographs and
then averaging them. Two methods of joint anlaysis are considered:

(a) Event concentration

(b) Event superposition

(a) Event Concentration: By this method, an average unit hydrograph for a catchment can be
derived from the resulting hypothetical event obtained by changing a number of events end
to end. This method makes use of the assumption of time invariance that in inherent in the
adoption of a unit hydrograph based model.

The principle time invariance states that the response to net rainfall is independent of the
time at which the rainfall occurred. Therefore, it is permissible to alter the time attached to
an event provided that the relative timing of net rainfall and quick response runoff is
preserved. A simple example of event concentration is shown in Fig. 21.

(b) Event Superposition: This technique relies on the unit hydrograph assumptions of linearity
and time-invariance. The steps to be carried out, for deriving the average unit hydrograph by

Workshop on Flood Management, January 18-22, 2010, National Institute of Hydrology, Roorkee-247667 L-7 36/40
 Flood Estimation by Unit Hydrograph Techniques

this method, are:

• Preserve the relative timing of net rainfall and quick response runoff for each event.
• Align the peak elements of net rainfall of each event.
• Superpose the event excess rainfall data by adding the corresponding excess rainfall
blocks of each event, obtained from the alignment of the peak rainfall of each event.
• Similarly superpose the surface runoff hydrographs of each event.
• Derive the unit hydrograph from the resulting superposed event. This gives the
average unit hydrograph for the catchment.

Fig. 22 illustrates the superposition for the same example as that used to illustrate concentration.

CATCHMENT SIZE LIMITATION FOR UNIT HYDROGRAPH

Areal uniformity is a principle of the unit hydrograph theory, and for larger catchments it is
difficult to obtain. The upper limit for the application of this principle is usually adopted as 2500
to 7500 sq. km. If the catchment is greater than 7500 sq. km., one unit hydrograph should not be

Workshop on Flood Management, January 18-22, 2010, National Institute of Hydrology, Roorkee-247667 L-7 37/40
 Flood Estimation by Unit Hydrograph Techniques

used, but the catchment should be divided into sub-areas usually major tributaries.

UH APPLICATION ON SMALL CATCHMENTS FOR FLOOD ESTIMATION

Computation of Direct Surface Runoff Hydrograph

The ordinates of the unit hydrograph must be multiplied by the depth of excess rainfall to
provide the different hydrographs of direct surface runoff, one for each of the intensities given in
the storm. These must then be superimposed with the correct time lag and added to give the
direct surface runoff hydrograph due to composite storm.

FLOOD ESTIMATION FOR LARGE CATCHMENTS

For the catchments having area more than 5000 Sq. Km., to apply the principles of Unit
Hydrograph, the whole catchment is sub-divided into a suitable number of sub-catchments and
the flood hydrograph for each sub-catchment is computed and the combined contributions from
each sub catchment are routed through the respective river reaches or reservoirs using an
appropriate flood routing technique to estimate the final flood for this large size catchment.
The following steps are involved in the derivation of flood hydrograph for the whole
catchment:
(i) Divide the whole catchment into suitable number of sub catchments.
(ii) Identify the reaches and calculate individual reach length. For each reach,
identify total no of sub-catchments at the start of the reach.
(iii) For each sub-catchment, collect either flood hydrograph or storm (design storm),
baseflow and unit hydrograph (design unit hydrograph) details.
(iv) Similarly for the reach, identify routing procedure and routing parameters. If
routing parameters are not available then first calculate them using inflow and
outflow hydrograph along with other information for that reach.
First step involves the calculation of flood (design flood) hydrograph for a sub-catchment
if storm details and Unit Hydrograph details are available. Then, flood hydrograph(s) of
sub-catchment(s) at the start of the first reach are routed through the reach to get the routed
outflow hydrograph. This outflow hydrograph combined with flood hydrograph of
sub-catchment(s) meeting at start of second reach is routed through second reach to get routed
flood hydrograph at the end of this reach. Likewise, the routed flood hydrograph may be
computed for the last reach which will be the required flood hydrograph for the whole catchment.

Workshop on Flood Management, January 18-22, 2010, National Institute of Hydrology, Roorkee-247667 L-7 38/40
 Flood Estimation by Unit Hydrograph Techniques

APPLICATION OF UNIT HYDROGRAPH

Having derived a unit hydrograph, the question arises as to what use do we put it? Some of the
current applications are:

• Estimation of design flood;

• Estimation of runoff from ungauged catchments
• Filling up the missing records
• Flood forecasting to estimate flood flows from rainfall records in real time.

BIBLIOGRAPHY

Chow, V.T., 1964. Runoff Section 14, Handbook of Applied Hydrology. McGraw Hill Book
Co., New York.
Kuchment, L.S., 1967. Solution of Inverse Problems for Linear Flow Models. Soviet
Hydrology, Selected Papers, Vol. 2, p.94.
Nash, J.E., 1982-83. Lecture Notes on Hydrologic System. International post graduate course in
Hydrology, University College, Galway, Ireland (Unpublished).
Singh, R.D. and S.M.Seth, 1984-85. Comparative Study of Unit Hydrograph Methods. Report
No CS-7, National Institute of Hydrology, Roorkee.
Singh, R.D., 1985. Application of an Efficient Smoothed Least Square Technique for Unit
Hydrograph Derivation. Hydrology Jour. of IAH, Vol. 8, No. 3, p.21.
Singh, R.D., 1987. Derivation of average Unit Hydrograph Using Smooth Least Square
Technique. Jour. of the Institution of Engrs. (India), Vol. 61, May.
U.S. Army Corps of Engineers, 1973. Hydrograph Anlaysis. Vol. 4, Oct.
Willson, E.M., 1969. Engineering Hydrology. Mac Millon, London.

Workshop on Flood Management, January 18-22, 2010, National Institute of Hydrology, Roorkee-247667 L-7 39/40
 Flood Estimation by Unit Hydrograph Techniques

(1) (2) (3) (4) (5) (6)= (7) (8) (9)

(5)-(4)
Time unit Rainfall Trial Unit Excludes Actual (6) ) Ui Adjusted
periods excess mm Hydrograph PEMAX hydrograph qi-3PEi Ui PEMAX (m3/s) for volume
(m3/s) of surface correctness
runoff U (m3/s)
U1 U2 U3 U4 (m3/s)
0 q0
1 PE1 PE1 U1 q1
2 q2
3 q3
4 q4
5 q5
6 q6

Workshop on Flood Management, January 18-22, 2010, National Institute of Hydrology, Roorkee-247667 L-7 40/40
 Flood Estimation by Unit Hydrograph Techniques

Time PE For 100 mm Trial (1)

Period (mm) unit hydrograph Unit Hydrograph (m3/s)
(6 hrs) excess rainfall
(mm) PE/100 U1 U2 U3 U4 U5 U6 U7
1125 1125 1125 1125 1125 1125 1125
1 40 0.40 450 450 250 - - -
2 100 1.00 - 450 450 1050 600 863 821
3 60 0.60 675 - 450 1125 2050 925 1025 975
4 675 - 450 1125 4350 3225 2175 2069
5 675 - 450 1125 4150 3025 2075 1974
6 675 - 450 1125 2300 1175 1150 1094
7 675 - 450 1125 1070 -55 535 509
8 675 - 675 450 -225 450 425
9 675 675 120 - - -
Total 200 15790 8273 7870

Workshop on Flood Management, January 18-22, 2010, National Institute of Hydrology, Roorkee-247667 L-7 41/40