Chapter 5

HYDROGRAPHS
5.1 HYDROGRAPH COMPONENTS
A hydrograph is a graph showing discharge (i.e., stream flow at the concentration point)
versus time. The various components of a natural hydrograph are shown in Fig. 5.1. At the
beginning, there is only base flow (i.e., the ground water contribution to the stream) gradually
depleting in an exponential form. After the storm commences, the initial losses like intercep-
tion and infiltration are met and then the surface flow begins. The hydrograph gradually rises
Effective duration
of net rain
intensity i (cm/hr)

Rainfall, P
Hyetograph
Rainfall

Net rain, Pnet

C.g of Pnet Pnet = P – Losses
f-curve
Infil-
tration loss (loss rate)
Time t (hr or min)
N days after peak
Basin lag
tp Peak flow 0.2
A
Crest N(days) = (empirical)
1.21
2
A = Area of basin (km )

Time of Inflection
Discharge Q (cumec)

concentration point
tc
Total runoff hydrograph (TRH)
of stream-flow = 1 + 2
Rising 1
limb Recession limb or
Direct surface curve (or falling limb)
runoff volume
= Pnet × A
Ground water
depletion curve extrapolated
Point
of rise Ground water
N days after peak
depletion curve
B.F. Separation line GW depletion curve
Elevated
2 Ground water contribution
ground water table contributes more
of stream flow (base flow B.F.)

Time t (hr or days)

Fig. 5.1 Components of streamflow hydrograph

117
118 HYDROLOGY

and reaches its peak value after a time tp (called lag time or basin lag) meaured from the
controid of the hyetograph of net rain. Thereafter it declines and there is a change of slope at
the inflection point, i.e., there has been, inflow of the rain up to this point and after this there
is gradual withdrawal of catchment storage. By this time the ground water table has been
built up by the infiltrating and percolating water, and now the ground water contributes more
into the stream flow than at the beginning of storm, but thereafter the GWT declines and the
hydrograph again goes on depleting in the exponential form called the ground water depletion
curve or the recession curve. If a second storm occurs now, again the hydrograph starts rising
till it reaches the new peak and then falls and the ground water recession begins, Fig. 5.2.
Storm 3
Storm 2
Intensity
i (cm/hr)

Storm 1
Pnet1 Pnet2 Pnet3
Loss Loss Loss

3 peaks due to
3-storms in
succession
Discharge Q (cumec)

Recession
lines

Estimated base flow

Time t (hr or days)

Fig. 5.2 Hydrograph with multiple peaks

Thus, in actual streams gauged, the hydrograph may have a single peak or multiple
peaks according to the complexity of storms. For flood analysis and derivation of unit
hydrograph, a single peaked hydrograph is preferred. A complex hydrograph, however, can be
resolved into simple hydrographs by drawing hypothetical recession lines as shown in Fig. 5.2.
It has been found from many hydrographs that the ground water depletion curves for a
given drainage basin are nearly the same and hence it is termed as the normal ground water
depletion curve. It has been found that such curves, or at least their segments, follow a simple
inverse exponential function of the elapsed time of the form.
Qt = Q0 Kr–t ...(5.1)
where Q0 = discharge at start of period
Qt = discharge at end of time t
Kr = recession constant
As Qt is the derivations of storage with respect to time,
– d St = Qt dt

C-9\N-HYDRO\HYD5-1.PM5 118
HYDROGRAPHS 119

Qt
St = –

ax
z Qt dt = – z (Q0 Kr–1) dt

∴
log e K r
3 ax = St =
log a
...(5.2) z
Hence, the discharge at any time is proportional to the water remaining in storage i.e.,
Q0 S0
= ...(5.2 a)
Qt St
Taking logarithms on both sides of Eq. (5.1)
log Qt = log Q0 – t log Kr
which is in the straight line form y = mx + c with y = log Qt, x = t, and m = – log Kr. The value
of Kr can be determined by plotting the recession data i.e., Qt vs. t on a semi-log paper, taking
care to select periods of little or no direct runoff; see Example 17.3.

a. Semi-log plot of continuous

Discharge Q (cumec)

discharge hydrograph
Log scale

a
c

e
b
g
d
f
h i

Time t (hr or days) (natural scale)

a b. Semi-log plot of
c master depletion curve
e
g Lower ends of curves
Log Q

b
d are tangential to a
Master i
f common line
depletion curve h
(semi-log plot) j

Time t (hr or days) (natural scale)

Discharge Q (cumec)

Master depletion
curve (natural scale)

c. Master depletion
curve drawn to linear scale

Time t (hr or days)

Fig. 5.3 Composite ground water depletion curve

C-9\N-HYDRO\HYD5-1.PM5 119
120 HYDROLOGY

If a continuous stream flow record is available for a number of years, the hydrograph
can be plotted on a semi-log paper, i.e., log Q vs. t, Fig. 5.3. Starting with the lowest recession
flow line, a line is drawn tangential to the lower portion on a tracing paper. This tangent line
is progressively extended by moving the tracing-paper towards the origin with the abscissae
coincident, such that the line is tangential to the lower portion of the successive depletion
curves of increasing magnitude. This common line is the log plot of the master depletion curve,
which is then converted to linear vertical scale and is called the composite ground water depletion
curve.
A composite ground water depletion curve can be constructed from the recession graphs
resulting from a number of storms. The various segments of the recession graphs are shifted
with respect to the time axis until they appear to match and then, an average or composite
curve is drawn through them as shown in Fig. 5.4.

20

18

16
Discharge Q (cumec 10 or 100)

Master depletion curve

14

12
Storm recession curves
10 moved (to match) w.r.t.
time-axis
8

0
3 6 9 12 15 18 21 24 27
Time t (hr or days)

Fig. 5.4 Composite depletion curve

5.2 SEPARATION OF STREAMFLOW COMPONENTS

Barnes (1940) proposed that the stream flow components may be separated by plotting the
hydrograph on a semi-log paper (Fig. 5.5). The tail end of the hydrograph plots as a straight
line, i.e., ground water recession (CD). If this straight line plot is extended backwards up to the
point E directly under the inflection point I and line BE drawn, the area under BEC represents
the ground water contribution to the stream flow. If the ordinates of this area are deducted
from the ordinates of the total hydrograph and replotted, the hydrograph of surface runoff and
interflow (subsurface flow) is obtained, which plots as a straight line (HG) at the tail end. By
extending this line backwards up to the point L directly under I and drawing the line FL, the
area under FLG gives the interflow component. By deducting the ordinates of this from the
ordinates of the hydrograph of surface runoff and interflow, the hydrograph of surface runoff
is replotted whose tail end again, may plot as a straight line representing the surface recession

C-9\N-HYDRO\HYD5-1.PM5 120
HYDROGRAPHS 121

or channel storage. The slopes of the straight line plots at the tail ends of the separated
hydrographs give the respective recession constants.
Example 5.1 The mean daily streamflow data from a drainage basin is given below. It is known
that the recession limb of the discharge hydrograph has components of channel storage, interflow
and base flow. Find the values of the recession coefficients for each of the three components.

Date Mean daily discharge Date Mean daily discharge

(cumec) (cumec)

1978, Oct. 4 278 1978, Oct. 14 179

5 265 15 167
6 5350 16 157
7 8150 17 147
8 6580 18 139
9 1540 19 131
10 505 20 123
11 280 21 117
12 219 22 111
13 195 23 105
24 100

Also determine
(a) ground water storage on October 14, 1978.
(b) ground water storage and stream flow on October 30, 1978, assuming no rainfall
during the period.
Solution The discharge hydrograph is drawn on a semi-log paper and the flow components
are separated by the method proposed by Bernes as shown in Fig. 5.5. The recession coeffi-
cients (Kr) for the three components of base flow (ground water contribution), interflow and
channel storage are computed as 1.059, 2.104 and 4.645, respectively.
(a) Ground water storage (S0) on October 14, 1978 when the ground water depletion
starts.
Q0 179 × 86400 m 3 /day
S0 = = = 2.75 × 108 m3
log e kr log e 1.06/day
(b) Stream flow on October 30, 1978, i.e., after 16 days
Qt = Q0 Kr–t
Q16 days = 179 (1.06)–16 = 71.6 cumec
Ground water storage (St) on October 30, 1978 can be determined from
Q0 S
= 0
Qt St
Qt
or St = S0 ×
Q0
716
.
∴ S16 days = (2.75 × 108) = 1.10 × 108 m3
179

C-9\N-HYDRO\HYD5-1.PM5 121
122 HYDROLOGY

10000
8 I
6 Semi-log paper

4 Total runoff hydrograph

3 (of stream flow)

–t
2 Qt = Q o K r
Log Qt = log Qo – t log Kr

þ
ý
ü

þ
ý
ü

þ
ý
ü
1000
8 y c + x m (Slope m is –ve)
6 m = – log Kr, Kr = daily depletion factor
Log-scale

L or recession constant
4
Surface runoff plus interflow
3 E
A B Ground water recession
C
2 log 330/100
–m = = 0.026
20 days = log K
r
D
100
–1
8 Kr = 1.06 days
Discharge Q (cumec)

6
5
4 Surface
runoff
3
Interflow
2
Surface
recession
10
H
8
6
5
K r = 4.6

Kr=

4
2.1

3
4 days

0d

2
ays
–1
–1

F M G
1
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Date Days – Natural scale

Fig. 5.5 Separation of streamflow components (after Barnes, 1940)

5.3 HYDROGRAPH SEPARATION

For the derivation of unit hydrograph, the base flow has to be separated from the total runoff
hydrograph (i.e., from the hydrograph of the gauged stream flow). Some of the well-known
base flow separation procedures are given below, Fig. 5.6.
(i) Simply by drawing a line AC tangential to both the limbs at their lower portion.
This method is very simple but is approximate and can be used only for preliminary estimates.
(ii) Extending the recession curve existing prior to the occurrence of the storm up to the
point D directly under the peak of the hydrograph and then drawing a straight line DE, where
E is a point on the hydrograph N days after the peak, and N (in days) is given by
N = 0.83 A0.2 ...(5.3)

C-9\N-HYDRO\HYD5-1.PM5 122
HYDROGRAPHS 123

Peak N days
B

I 0.2
Discharge Q (cumec)

N (days) = 0.84 A (empirical)

2
A = Drainage area (km )

Direct
surface Four separation procedures
runoff

Point of F
rise E (N days after peak)
A 4 3 G Recession
2 C
1
2 Ground water contribution
D to stream (base flow)
Time t (hr or days)

Fig. 5.6 Hydrograph separation

where A = area of the drainage basin, km2 and the size of the areas of the drainage basin as a
guide for the values of N are given below :

Area of drainage basin (km2) 250 1250 5000 12500 25000

Time after peak, N (days) 2 3 4 5 6

(iii) Simply by drawing a straight line AE, from the point of rise to the point E, on the
hydrograph, N days after the peak.
(iv) Construct a line AFG by projecting backwards the ground water recession curve
after the storm, to a point F directly under the inflection point of the falling limb and sketch an
arbitrary rising line from the point of rise of the hydrograph to connect with the projected base
flow recession. This type of separation is preferred where the ground water storage is rela-
tively large and reaches the stream fairly repidly, as in lime-stone terrains.
Many a time a straight line AE meets the requirements for practical purposes. Location
of the point E is where the slope of the recession curve changes abruptly, and as a rough guide
E is N days after the peak. In all the above four separation procedures, the area below the line
constructed represents the base flow, i.e., the ground water contribution to stream flow. Any
further refinement in the base flow separation procedure may not be needed, since the base
flow forms a very insignificant part of high floods. In fact, very often, a constant value of base
flow is assumed.
In Fig. 5.7, ADE is the assumed base flow. Actually, when the water level in the stream
rises due to floods, the stream feeds the ground water and the permeable boundaries (called
bank storage, Fig. 5.8) and this is termed as negative base flow. But actually the variation in
base flow may be much less than indicated in Fig. 5.7 depending on the permeability of the
boundaries and the gradient of the phreatic surface. The resulting variation will be a slight dip
from the extrapolated depletion curve, followed by a gradual rise to a higher level than the
initial value as the flood recedes as shown in Fig. 5.9.

C-9\N-HYDRO\HYD5-1.PM5 123
124 HYDROLOGY

Discharge Q (cumec)
Discharge
hydrograph
Surface
runoff
Recession
curves E
Increase in base flow
A after flood

Negative base flow

Time t (hr) D (bank storage)

Fig. 5.7 Negative baseflow

Bank storage
GW
T
High water level Permeable
banks of stream

Low water level GWT

Ground water flow
Ground water flow
(during low water period)

Fig. 5.8 Bank storage

Discharge Q (cumec)

Discharge
hydrograph
Direct
surface
runoff
Separation line
Recession
curve
Base flow Extrapolated
depletion curve
Time t (hr)

Fig. 5.9 Actual baseflow variation

5.4 UNIT HYDROGRAPH

The unit hydrograph is defined as the hydrograph of storm runoff resulting from an isolated
rainfall of some unit duration occurring uniformly over the entire area of the catchment, pro-
duces a unit volume (i.e., 1 cm) of runoff.
Derivation of the unit hydrographs. The following steps are adopted to derive a unit
hydrogaph from an observed flood hydrograph (Fig. 5.10).

C-9\N-HYDRO\HYD5-1.PM5 124
HYDROGRAPHS 125

intensity i (cm/hr) Rainfall, P

Pnet = 26.6 cm = Runoff
Rainfall

c.g. of Pnet
f-index
Time t (min)
tr = 3 hr = Effective duration of storm
300 1
tp = 10 hr peak flow 0.2 3
2 N = 0.84 (40) = 1 day (too long)
Basin lag 4
3
SQd.t 986 (3 × 60 × 60) m
250 Pnet = =
A 6 2
40 × 10 m
Total
Discharge (Stream flow) Q (cumec)

= 0.266 m
run off or 26.6 cm ... 1
200 hydrograph
SQd = SDRO = 986 cumec

150 Pnet = P – Losses

Direct surface = S(i – )t ... 2
runoff = 26.6 cm
t = duration of rain (min)
100 120 = TRO = Q for which i > f P from 1 & 2
net
75 = DRO = TRO – BFO should balance
Point of = Qd = 120 – 45 DSR stops recession
rise [Link] e of separation Recession
50 starts Arbitrary lin
BFO Base flow
47 46 45 45 45 46 48 50 53 54 57 60 55 51 50
0
2 5 8 11 14 17 20 23 2 5 8 11 14 17 20 23
1–3–1970 2–3–1970
Pnet =
75/26.6 = 2.82

1 cm Date and time hr

10 Peak of UG
Discharge – cumec

tr tp TRO – BFO = DRO

8 2 Unit
hydrograph DRO = UGO (for which P = 1 cm)
tr Pnet net
6
Area under tr = 3 hr = unit hydrograph duration
curve = 1 cm
4 of runoff tr
DRO Time to peak = tp +
2 2.82 = UGO = 2
Pnet
0
0 3 6 9 12 15 18 21 24 27 30 33
Hours from beginning of runoff
Runoff duration T = 32 hr

Fig. 5.10 Derivation of a unit hydrograph (Example 5.1)

(i) Select from the records isolated (single-peaked) intense storms, which occurring
uniformly over the catchment have produced flood hydrographs with appreciable runoff (>1
cm, say, 8 to 16 cm). The unit period selected should be such that the excess rainfall (i.e., Pnet)
occurs fairly uniformly over the entire drainage basin. Larger unit periods are required for
larger basins. The unit periods may be in the range of 15-30% of the ‘peak time’ period, i.e., the
time from the beginning of surface runoff to the peak, and the typical unit periods may be 3, 6,
8, 12 hours. (The time of concentration may be a little longer than the peak time).

C-9\N-HYDRO\HYD5-1.PM5 125
126 HYDROLOGY

The unit storm is a storm of such duration that the period of surface runoff is not much
less for any other storm of shorter duration.
(ii) Select a flood hydrograph, which has resulted from a unit storm chosen in item (i)
above.
(iii) Separate the base flow from the total runoff (by the well-known base flow separa-
tion procedures).
(iv) From the ordinates of the total runoff hydrograph (at regular time intervals) deduct
the corresponding ordinates of base flow, to obtain the ordinates of direct runoff.
(v) Divide the volume of direct runoff by the area of the drainage basin to obtain the net
precipitation depth over the basin.
(vi) Divide each of the ordinates of direct runoff by the net precipitation depth to obtain
the ordinates of the unit hydrograph.
(vii) Plot the ordinates of the unit hydrograph against time since the beginning of direct
runoff. This will give the unit hydrograph for the basin, for the duration of the unit storm
(producing the flood hydrograph) selected in item (i) above.
In unit hydrograph derivation, such storms should be selected for which reliable rain-
fall and runoff data are available. The net rain graph (hyetograph of excess rain) should be
determined by deducting the storm loss and adjusting such that the total volume of net storm
rain is equal to the total volume of direct surface runoff. The unit hydrograph derived, which,
when applied to the known net rain data, should yield the known direct runoff hydrograph.
The steps given above for the derivation of unit hydrograph can be formulated as fol-
lows (exemplified in Fig. 5.10).
Pnet = P – Losses ...(5.4)
ΣQd t
or Pnet = , Qd = DRO ...(5.4 a)
A
TRO—BFO = DRO ...(5.4 b)
DRO
= UGO ...(5.4 c)
Pnet
where P = total rainfall
Pnet = net precipitation (from hyetograph) or direct runoff as equivalent depth
over the basin.
Losses = due to infiltration (Fp), etc.
A = area of the drainage basin
Qd, DRO = direct runoff ordinate
TRO = total runoff ordinate
t = time interval between successive direct runoff ordinates
BFO = base flow ordinate
Elements of unit hydrograph. The various elements of a unit hydrograph are shown in
Fig. 5.11.
Base width (T)—The period of direct surface runoff (due to a unit storm) of the unit
hydrograph is called the time base or the base width.

C-9\N-HYDRO\HYD5-1.PM5 126
HYDROGRAPHS 127

Unit

intensity i
1
period i= cm/hr

Rainfall
tr
tr
Pnet = 1 cm = i × tr

Recession time Tr
Lag time
tp
Qp tr/2 Peak of UG, Qp cumec

Time to peak
Concen-
Discharge Q (cumec)

tration time I, inflection

tc point

Unit
hydrograph
Area under
curve Channel
º 1 cm of storage
runoff over
catchment
area A

Time base T (hr)

(Runoff duration)
Time t (hr)

Fig. 5.11 Elements of unit hydrograph

Unit storm—The strom of unit duration (i.e., duration of the unit hydrograph) regard-
less of its intensity is called unit storm.
Unit period—The time duration of the unit storm (i.e., the duration of the unit
hydrograph) is called unit period.
Lag time (tp)—The time from the centre of a unit storm to the peak discharge of the
corresponding unit hydrograph is called lag time.
Recession time (Tr)—The duration of the direct surface runoff after the end of the excess
or net rainfall, is called recession time in hydrograph analysis.

Propositions of the Unit Hydrograph

The following are the basic propositions of the unit hydrograph:
(i) Same runoff duration. For all unit storms of different intensities, the period of sur-
face runoff (i.e., time base, base width or base period) is approximately the same, although
they produce different runoff volumes (Fig. 5.12 (a)).
(ii) Proportional ordinates. For unit storms of different intensities, the ordinates of the
hydrograph at any given time, are in the same proportion as the rainfall intensities
(Fig. 5.12 (a)).
(iii) Principle of superposition. If there is a continuous storm and/or isolated storms of
uniform intensity net rain, they may be divided into unit storms and hydrographs of runoff for
each storm obtained, and the ordinates added with the appropriate time lag to get the com-
bined hydrograph (Fig. 5.12 (b)).

C-9\N-HYDRO\HYD5-1.PM5 127
128 HYDROLOGY

Intensity
(cm/hr)
i = 1 , xi = x cm/hr 1 × t = 1 cm
tr tr r
x cm ü tr
xi ýP x × t = x cm
1 cm þ net tr r
i

tr
x Qp
Hydrograph (produces
Discharge Q (cumec)

x cm of runoff)
Run off
= x cm Peaks occur
Qp at same time
x Qy Proportional ordinates
tr – hr UG (1-cm runoff)
Qy Qp
Run off Same time base T
Qy = 1 cm

t1 tpeak
Time base T

Time t (hr)
(a) Proportional ordinates and same time base

tR = 2 t r Continuous storm split into two unit storms

intensity i
Rainfall

y cm
x cm z cm
2
1 3
Pnet due to 3-storms
tr tr tr unit period tr – hr
Combined hydrograph (obtained
by summing the ordinates of the
three component hydrographs: 1, 2 & 3)
Discharge Q (cumec)

UGO × y

1 3 Hydrographs
2
corresponding
to storm 1, 2 & 3
tr – hr UGO UGO × z

UGO × x

Time t (hr)
(b) Principle of superposition

Fig. 5.12 Propositions of unit hydrograph

(iv) Same distribution percentages. If the total period of surface runoff (i.e., time base or
base width) is divided into equal time intervals the percentage of surface runoff that occurs
during each of these periods will be same for all unit storms of different intensities (Fig. 5.13).

C-9\N-HYDRO\HYD5-1.PM5 128
HYDROGRAPHS 129

intensity i (cm/hr)
y
y cm = i =cm/hr
Pnet tr S Mid–ordinates of unit periods
x = 0.5 + 2.0 + 4.7 + 7.1 + 5.6
x cm = i = cm/hr
Rainfall

tr + 3.5 + 2.5 + 1.6 + 0.9 + 0.3

1 = 28.7 units
1 cm = i = cm/hr
tr
% of runoff in unit period 1
0.5
= × 100 = 1.74 » 2 (rounded)
Unit period 28.7
12 and like that.
25% runoff
11 occurs
in unit 4 UGO × y = y cm runoff hydrograph
Discharge Q (cumec) (say, 10 cumec units)

10 period 4 25%

9 3 5
16% 19.5
% UGO × x = x cm runoff hydrograph
8
25%
7 + ve & – ve Distribution percentages
areas (Same for all units storms)
6 should 19.5%
balance + UGO = 1 cm runoff unit hydrograph
5 16%
–
4 Area under UG = Area under bar graph
7.1 units 12%
3 UGO Mid-ordinates of unit periods
5.6 9%
7% 4.7 (1 cm)
2 3.5 5.5% Bernard’s distribution graph

2.0 units 2.5 3%

1 2% 1.6
0.9 1%
0.5
0
1 2 3 4 5 6 7 8 9 10 If tr = 4 hr, T = 40 hr
Time base T divided into 10 unit periods
Time t (hr)

Fig. 5.13 Distribution percentages same for all unit storms

Limitation of the Unit Hydrograph. Certain limitations are inherent in the unit
hydrograph theory. The runoff hydrograph reflects the combined effects of rainfall factors,
loss factors and physiographic factors. The design storm continuing for several unit periods
may not have the same areal distribution for each time increment. Storm movements also
affect the proportions of the unit hydrograph if the basin is large. Hence, the unit hydrograph
can not be applied for basins larger than 5000 km2. For basins larger than 5000 km2, unit
hydrographs for the principal sub-areas or sub-basins are developed and the hydrographs of
runoff determined for each sub-area. These hydrographs are then combined, through flood
routing procedure, to get the resulting hydrograph at the required section.

Derivation of Unit Hydrograph

Example 5.1 (a) The runoff data at a stream gauging station for a flood are given below. The
drainage area is 40 km2. The duration of rainfall is 3 hours. Derive the 3-hour unit hydrograph
for the basin and plot the same.

C-9\N-HYDRO\HYD5-1.PM5 129
130 HYDROLOGY

Date Time Discharge Remarks

(hr) (cumec)

1-3-1970 2 50
5 47
8 75
11 120
14 225
17 290 ← Peak
20 270
23 145
2-3-1970 2 110
5 90
8 80
11 70
14 60
17 55
20 51
23 50

State the peak of the unit hydrograph you derive.

Solution See Table 5.1.

5.5 UNIT HYDROGRAPH FROM COMPLEX STORMS

Unit hydrographs from complex storms, involving varying intensities of rain can be obtained
by considering the complex storm as successive unit storms of different intensities and the
runoff hydrograph (due to complex storm) as the result of superposition of the successive
storm hydrographs. The ordinates of each storm hydrograph are obtained as ‘the storm inten-
sity times the corresponding ordinate of the unit hydrograph’ as shown in Fig. 5.14. The unit
hydrograph ordinates u1, u2, ... are thus obtained by writing a series of equations for each of
the ordinates Q1, Q2, .... of the runoff hydrograph (due to complex storm) and successively
solving them. In Fig. 5.14,
Q1 = xu1 ∴ u1 = ?
Q2 = xu2 + yu1 ∴ u2 = ?
Q3 = xu3 + yu2 + zu1 ∴ u3 = ?
and so on. Thus, the tr–hour unit graph ordinates can be determined. Although the method is
straight forward, errors will creep in due to the assumptions on the intensity and duration of
rainfall and deduction of an assumed base flow; many trials are required to get a reasonable
unit graph.

C-9\N-HYDRO\HYD5-1.PM5 130
HYDROGRAPHS 131

Table 5.1 Derivation of the 3-hour unit hydrograph. Example 5.1

Date Time TRO1 BFO2 DRO2 UGO4 Time5

(hr) (cumec) (cumec) (3)–(4) (5) ÷ Pnet from begin-
(cumec) (cumec) ning of
surface run
off (hr)

1 2 3 4 5 6 7

1-3-1970 2 50 50 — — —
5 47 47 0 0 0
8 75 46 29 1.09 3
11 120 45 75 2.82 6
14 225 45 180 6.77 9
17 290 45 245 9.23 12
20 270 46 224 8.44 15
23 145 48 97 3.65 18
2-3-1970 2 110 50 60 2.26 21
5 90 53 37 1.39 24
8 80 54 26 0.98 27
11 70 57 13 0.49 30
14 60 60 0 0 33
17 55 55 —
20 51 51 —
23 50 50 —
Σ DRO = 986 cumec

1TRO—Total runoff ordinate = gauged discharge of stream

2BFO—Base flow ordiante read from graph separation line shown in Fig. 5.10 (a). N = 0.83 A0.2
0.2
= 0.89(40) = 1.73 days = 1.73 × 24 = 41.4 hr from peak, which is seen not applicable here; hence an
arbitrary separation line is sketched.
3DRO—Direct runoff ordinate = TRO—B.F.O.
4UGO—Unit hydrograph ordinate

DRO ΣDRO . t 986(3 × 60 × 60)m 3

= ; Pnet = =
Pnet A 40 × 10 6 m 2
= 0.266 m = 26.6 cm
5Time from beginning of direct surface runoff is at 5 hr on 1-3-1970, which is reckoned 0 hr for
unit hydrograph. The time base for unit hydrograph is 33 hours.
The 3-hour unit hydrogryph is plotted in Fig. 5.10 (b) to a different vertical scale and its peak is
9.23 cumec.

C-9\N-HYDRO\HYD5-1.PM5 131
132 HYDROLOGY

intensity i (cm/hr)
Rainfall

Pnets due to 3-storms

y cm
x cm z cm
tr tr tr
Q3

Q2

z u1
Observed hydrograph
Discharge Q (cumec)

Q1 = x u1
y u1

Q2 = x u2 + y u1
Q3 = x u3 + y u2 + z u1, etc
Q4
Successively solve
for u1, u2, u3, etc.
z u2
y u2

which give tr – hr UGO

Q5
Q1 tr – hr
UGO
z u3
x u2

y u3

Q6 x u3
x u1

UG
y u4

Q7 u1 u2 z u4
x u4

u3 us y u
u4 x u5 5 0 z u5
0
0 0
Time t (hr)

Fig. 5.14 Derivation of unit hydrograph from multi-period storms

Example 5.2 The stream flows due to three successive storms of 2.9, 4.9 and 3.9 cm of 6 hours
duration each on a basin are given below. The area of the basin is 118.8 km2. Assuming a
constant base flow of 20 cumec, derive a 6-hour unit hydrograph for the basin. An average
storm loss of 0.15 cm/hr can be assumed.

Time (hr): 0 3 6 9 12 15 18 21 24 27 30 33
Flow (cumec): 20 50 92 140 199 202 204 144 84.5 45.5 29 20

Solution Let the 6-hour unit hydrograph ordinates be u0, u1, u2, u3, u4, ...., u7 at 0, 3, 6, 12, ....,
21 hours, respectively. The direct runoff ordinates due to the three successive storms (of 6
hours duration each) are obtained by deducting the base of flow of 20 cumec from the stream-
flows at the corresponding time intervals as shown in Table 5.2. The net storm rains are ob-
tained by deducting the average storm loss as
0-6 hr: x = 2.9 – 0.15 × 6 = 2 cm
6-12 hr: y = 4.9 – 0.15 × 6 = 4 cm
12-18 hr: z = 3.9 – 0.15 × 6 = 3 cm
The equations can be easily arrived by entering in a tabular column and successively
solving them. The 6-hr unit hydrograph ordinates are obtained in the last column; of course
the ordinates are at 3-hr intervals since the streamflows are recorded at 3-hr intervals. The

C-9\N-HYDRO\HYD5-1.PM5 132
 Table 5.2 Derivation of a 6-hour unit hydrograph from a complex storm, Example 5.2.

HYDROGRAPHS

DRO due to** Equation

Time UGO* 1st storm 2nd storm 3rd storm Total DRO = TRO— BFO Solution
(hr) UGO × x UGO × y UGO × z 6-hr UGO

C-9\N-HYDRO\HYD5-1.PM5
0 u0 = 0 0 — — 0 = 20 – 20 u0 = 0

133
3 u1 2u1 — — 2u1 = 50 – 20 u1 = 15
6 u2 2u2 0 — 2u2 = 92 – 20 u2 = 36
9 u3 2u3 4u1 — 2u3 + 4u1 = 140 – 20 u3 = 30
12 u4 2u4 4u2 0 2u4 + 4u2 + 0 = 199 – 20 u4 = 17.5
15 u5 2u5 4u3 3u1 2u5 + 4u3 + 3u1 = 202 – 20 u5 = 8.5
18 u6 2u6 4u4 3u2 2u6 + 4u4 + 2u2 = 204 – 20 u6 = 3
21 u7 2u7 4u5 3n3 2u7 + 4u5 + 3u3 = 144 – 20 u7 = 0
24 4u6 3u4 4u6 + 3u4 = 84.5 – 20 Σ u = 110
27 4u7 3u5 4u7 + 3u5 = 45.5 – 20 ∴ u5 = 8.5 check for
30 3u6 3u6 = 29 – 20 ∴ u6 = 3 UGO derived
|VU
33 3u7 3u7 = 20 – 20 ∴ u7 = 0 W| above

*Except the first column, all other columns are in cumec.

**x = 2 cm, y = 4 cm, z = 3 cm
133
134 HYDROLOGY

last four equations in the table serve to check some of the UGO’s derived. Another check for
the UGO’s derived is that the area under the UG should give a runoff volume equivalent to
1 cm, i.e.,
Σu t
= 1 cm, in consistent units
A
Σu = sum of the UGO’s = 110 cumec
110 (3 × 60 × 60)
∴ = 0.01 m, or 1 cm
118.8 × 10 6
Hence, the UGO’s derived are correct and is plotted in Fig. 5.15 (b).
The UGO’s can also be derived by the method of least squares (Snyder, 1955) by writing
the direct runoff ordinate (Q) as
Q = a + b1x1 + b2x2 + b3x3 + ... + bixi ...(5.5)
where a = 0 theoretically, bi = ui , and x1, x2, x3,... are the rainfall excesses (Pnet) in successive
periods. In the least square technique, data from a number of flood events, for which values of
Qi and xi are established, are used to develop a set of average values of bi (= ui) and the unit
hydrograph derived. The method is more elegant but laborious.

Pnet due to 3 storms

Intensity i
(cm/hr)

4 cm 3 cm
2 cm
6 hr 6 hr 6 hr
200 220

180 200
3 u1

a. observed
160 180
hydrograph due
to 18-hr storm
140 160
Surface runoff Q (cumec)

3 u2

(u’s = 6-hr UGO)

Stream flow Q (cumec)

120 140
4 u2

100 120
4 u1

4 u3

80 100
3 u3

UGO × 4
4 u4

60 80
2 u2

40 60
3 u4

UGO × 3
2 u3

UGO × 2
2 u4

3 u6
4 u5

20 40
2 u1

3 u5
4 u5
2 u5

2 u6
0 20
Constant base flow = 20 cumec
0
0 3 6 9 12 15 18 21 24 27 30 33
Time t (hr)

Fig. 5.15 (a) Derivation of 6-hr UG from 18-hr complex storm hydrograph (Example 5.2)

C-9\N-HYDRO\HYD5-1.PM5 134
HYDROGRAPHS 135

40

35

30

Discharge Q (cumec)
6-hr UG derived
25
(u's = UGO)
20

u2

u3
15

10
u1

u4
5

us
u6
0
0 3 6 9 12 15 18 21
Time t (hr)

Fig. 5.15 (b) 6-hr unit hydrograph (derived) (Example 5.2)

Matrix method for unit hydrograph derivation from complex storm

In matrix notation
PU = Q

LM p 1 0 0 0 0 0 OP LMu OP1 LMq OP1

MM pp
2
3
p1
p2
0
p1
0
0
0
0
0
0 PP MMuu PP
2
3
MMqq PP
2
3
where P = U= Q=
MM00 p3
0
p2
p3
p1
p2
0
0
0
0
PP MMu0 PP
4 MMqq PP
4
5
MN0 0 0 p3 0 0 PQ MN0 PQ MNq PQ
6

By deducting base flow from the stream flow hydrograph Q is known, and with the
known rainfall excess P, the unit hydrograph U can be computed as
U = P–1 Q
The inverse for the precipitation matrix (P–1) exists only if it is a square matrix with a
non-vanishing determinant. However, by the use of the transpose of the precipitation matrix,
(PT ), a square matrix is obtained as
PT PU = PTQ
and the unit hydrograph matrix is obtained as
U = (PT P)–1 PTQ
Usually, the solution of the unit hydrograph matrix is performed on a digital computer
(Newton, 1967). Most computer installations have packaged programmes to work with large
systems and for determining the inverse and transpose of matrices.
Also, the number of streamflow hydrograph ordinates (n) is given by
n=j+i–1
where j = number of UGO
i = number of periods of rainfall

C-9\N-HYDRO\HYD5-1.PM5 135
136 HYDROLOGY

In Example 5.2, n = 12 ordinates at 3-hr interval (DRO of the first and last being zero) ;
i = 6 periods of 3-hr interval (= 3 × 2), then
n = j + i – 1; 12 = j + 6 – 1
∴ j = 7 UGO
Thus, the unit hydrograph consists of 7 ordinates at 3-hr intervals (of course, u0 = 0, and
u7 = 0).
As an assignment, the student is advised first to become well versed with the matrix
algebra and then apply the matrix method to derive the 6-hr UGO at 3-hr interval from the
streamflow data given in Example 5.2.
Average Unit Hydrograph. It is better if several unit graphs are derived for different
isolated (single peaked) uniform intensity storms. If the durations of storms are different, the
unit hydrographs may be altered to the same duration (sect .... or art .....). From several unit
hydrographs for the same duration, so obtained, an average unit hydrograph can be sketched
by computing the average of the peak flows and times to peak and sketching a median line,
(Fig. 5.16), so that the area under the graph is equal to a runoff volume of 1 cm.

Pnet = 1 cm

tr Average peak given by:

Qp3 Qp1 + Qp2 + Qp3
Qp =
Qp 3
Qp1
Qp2 t1 + t 2 + t 3
tpeak =
3 3

1 Median line sketched so that area

2 under curve = 1 cm runoff, gives tr-hr
average unit hydrograph
Discharge Q (cumec)

1 , 2 & 3 : tr-hr UG's

obtained from different
duration storms

t1
tpeak
t2
t3

Time base T of ave. UG

Time t (hr)

Fig. 5.16 Average unit hydrograph

Alteration of Unit Hydrograph Duration. It becomes necessary, in computation of

flood hydrographs, that the duration of the unit graph available should be altered to suit the
duration of the design storm (to be used for obtaining the flood hydrograph). Two cases arise:

C-9\N-HYDRO\HYD5-1.PM5 136
HYDROGRAPHS 137

Case (i) Changing a Short Duration Unit Hydrograph to Longer Duration. If the desired
long duration of the unit graph is an even multiple of the short, say a 3-hour unit graph is
given and a 6-hour unit graph is required. Assume two consecutive unit storms, producing a
net rain of 1 cm each. Draw the two unit hydrographs, the second unit graph being lagged by
3 hours. Draw now the combined hydrograph by superposition. This combined hydrograph
will now produce 2 cm in 6 hours. To obtain the 6-hour unit graph divide the ordinates of the
combined hydrograph by 2, Fig. 5.17. It can be observed that this 6-hour unit graph derived
has a longer time base by 3 hours than the 3-hour unit graph, because of a lower intensity
storm for a longer time.
Example 5.3 The following are the ordinates of a 3-hour unit hydrgraph. Derive the ordinates
of a 6-hour unit hydrograph and plot the same.

Time (hr) 3-hr UGO Time (hr) 3-hr UGO

(cumec) (cumec)

0 0 15 9.4
3 1.5 18 4.6
6 4.5 21 2.3
9 8.6 24 0.8
12 12.0

Solution
Table 5.3 Derivation of the 6-hour unit hydrograph. (Example 5.3)

3-hr 3-hr Total 6-hr

Time UGO UGO (2) + (3) UGO
(hr) (cumec) (logged) (cumec) (4) ÷ 2
(cumec) (cumec)

1 2 3 4 5

0 0 0 0
3 1.5 0 1.5 0.7
6 4.5 1.5 6.0 3.0
9 8.6 4.5 13.1 6.5
12 12.0 8.6 20.6 10.3
15 9.4 12.0 21.4 10.7
18 4.6 9.4 14.0 7.0
21 2.3 4.6 6.9 3.4
24 0.8 2.3 3.1 1.5
27 0.8 0.8 0.4

C-9\N-HYDRO\HYD5-1.PM5 137
138 HYDROLOGY

The 6-hour unit graph derived as above is shown in Fig. 5.17.

Intensity i
(cm/hr)
1 cm 1 cm, Pnet
2 unit storms

3 hr 3 hr
22

20

18
Combined hydrograph
16
(2 cm in 6 hr)
Discharge Q (cumec)

14
3-hr UG's lagged by 3 hr
12

10
6-hr UG (1 cm in 6 hr)
(comb. hyd. ord. ¸ 2)
8

0
0 3 6 9 12 15 18 21 24 27
Time t (hr)

Fig. 5.17 Changing a short duration UG to a long multiple duration

Case (ii) Changing a Long Duration Unit Hydrograph to a Shorter Duration by S-curve
Technique

5.6 S-CURVE METHOD

S-curve or the summation curve is the hydrograph of direct surface discharge that
would result from a continuous succession of unit storms producing 1 cm in tr–hr (Fig. 5.18). If
the time base of the unit hydrograph is T hr, it reaches constant outflow (Qe) at T hr, since 1 cm
of net rain on the catchment is being supplied and removed every tr hour and only T/tr unit
graphs are necessary to produce an S-curve and develop constant outflow given by,
2.78 A
Qe = ...(5.6)
tr
where Qe = constant outflow (cumec)
tr = duration of the unit graph (hr)
A = area of the basin (km2)
Given a tr-hour unit graph, to derive a tr′-hour unit graph (tr′ ≥ tr)—Shift the S-curve by
the required duration tr′ along the time axis. The graphical difference between the ordinates of
the two S-curves, i.e., the shaded area in Fig. 5.18 represents the runoff due to tr′ hours rain at

C-9\N-HYDRO\HYD5-2.PM5 138
HYDROGRAPHS 139

an intensity of 1/tr cm/hr, i.e., runoff of tr′/tr cm in tr′ hours. To obtain a runoff of 1 cm in tr′
hours (i.e., tr′-hour UG), multiply the ordinates of the S-curve difference by tr/tr′. This tech-
nique may be used to alter the duration of the given unit hydrograph to a shorter or longer
duration. The longer duration need not necessarily be a multiple of short.
i = 1 (cm/hr)
Intensity

Successive unit storms of Pnet = 1 cm

tr

1 cm 1 cm 1 cm 1 cm 1 cm 1 cm
To obtain tr – hr UG
tr tr tr tr tr tr
480 multiply the S-curve
difference by tr/tr¢
440
400 S-curve A
hydrograph
360
Lagged
320 S-curve
Discharge Q (cumec)

(lagged by Area
280 s-curve Constant outflow 2
tr – hr) difference
240 Unit hydrographs in 2.78 A(km )
= 1´ tr cm
Qe =
tr succession produce tr (hr)
200 constant outflow, Qe cumec
in tr – hr Cumec
160 B
tr – hr UG
120
80
40
2 4
0 tr 8 10 12 14 16 18 20 22 24
0 Lagged 6
Time t (hr)

Fig. 5.18 Changing the duration of UG by S-curve technique (Example 5.5)

Example 5.4 The ordinates of a 4-hour unit hydrograph for a particular basin are given be-
low. Derive the ordinates of (i) the S-curve hydrograph, and (ii) the 2-hour unit hydrograph,
and plot them, area of the basin is 630 km2.

Time Discharge Time Discharge

(hr) (cumec) (hr) (cumec)

0 0 14 70
2 25 16 30
4 100 18 20
6 160 20 6
8 190 22 1.5
10 170 24 0
12 110

Solution See Table 5.4

C-9\N-HYDRO\HYD5-2.PM5 139
 Table 5.4 Derivation of the S-curve and 2-hour unit hydrographs. (Example 5.4)

Time 4-hr UGO S-curve additions S-curve lagged S-curve 2-hr UGO 140
(hr) (cumec) (cumec) ordinates S-curve difference (6) × 4/3
(unit storms after every 4 hr = tr) (cumec) (cumec) (cumec) (cumec)
(2) + (3) (4) – (5)

1 2 3 4 5 6 7

C-9\N-HYDRO\HYD5-2.PM5
0 0 — — — — — 0 — 0 0
2 25 — — — — — 25 0 25 50

140
4 100 0 — — — — 100 25 75 150
6 160 25 — — — — 185 100 85 170
8 190 100 0 — — — 290 185 105 210
10 170 160 25 — — — 335 290 65 130
12 110 190 100 0 — — 400 355 45 90
14 70 170 160 25 — — 425 400 25 50
16 30 110 190 100 0 — 430 425 5 10*
18 20 70 170 160 25 — 445 430 15 30
20 6 30 110 190 100 0 436 445 –9 – 18*
22 1.5 20 70 170 160 25 446.5 436 10.5 21
24 0 6 30 110 190 100 436 446.6 – 10.5 – 21*

*Slight adjustment is required to the tail of the 2-hour unit hydrograph.

Col (5): lagged S-curve is the same as col (4) but lagged by tr′ = 2 hr.
tr
Col (7): col (6) × , t = 4 hr. tr′ = 2 hr.
tr ′ r
Col (3): No. of unit storms in succession = T/tr = 24/4 = 6, to produce a constant outflow.
2.78 A 2.78 × 630
Qe = = = 437 cumec, which agrees very well with the tabulated S-curve terminal value of 436. The S-curve
tr 4
additions can be written in one column, without having to write in 5 columns successively lagged by 4 hours (= tr), as is illustrated in the
example 5.5.
Plot col (4) versus col (1) to get the S-curve hydrograph, and col (7) versus col (1) to get the 2-hour unit hydrograph, as shown in
HYDROLOGY

Fig. 5.19.
HYDROGRAPHS 141

Example 5.5 The ordinates of a 4-hour unit hydrograph for a particular basin are given be-
low. Determine the ordinates of the S-curve hydrograph and therefrom the ordinates of the 6-
hour unit hydrograph.

Time 4-hr UGO Time 4-hr UGO

(hr) (cumec) (hr) (cumec)

0 0 12 110
2 25 14 70
4 100 16 30
6 160 18 20
8 190 20 6
10 170 22 1.5
24 0

Solution See Table 5.5.

Table 5.5 Derivation of the 6–hour UG for the basin (Example 5.5)

Time 4-hour S-curve1 S-curve lagged2 S-curve 6-hr

(hr) UGO additions ordinates C-curve difference UGO
(cumec) (cumec) (cumec) (cumec) (cumec) (cumec)
(2) + (3) (4) – (5) (6) × 4/6

1 2 3 4 5 6 7

0 0 → — 0 — 0 0
2 25 → — 25 — 25 16.7
4 100 + → 0 100 — 100 66.7
6 160 + → 25 185 0 185 123.3
8 190 100 290 25 265 176.7
10 170 185 355 100 255 170.0
12 110 290 400 185 215 143.3
14 70 355 425 290 135 90.0
16 30 400 430 355 75 50.0
18 20 425 445 400 45 30.0
20 6 430 436 425 11 7.3
22 1.5 445 446.6 430 16.5 11.0
24 0 436 436 445 –9 – 6.0

1—Start the operation shown with 0 cumec after tr = 4 hr.

2—Lag the S–curve ordinates by tr′ = 6 hr.
Plot col (4) vs. col (1) to get S-curve hydrograph and col (7) vs. col (1) to get 6-hr unit hydrograph
as shown in Fig. 5.19.

C-9\N-HYDRO\HYD5-2.PM5 141