Professional Documents
Culture Documents
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
Rainfall, P
Rainfall
Hyetograph
intensity i
Time of Inflection
concentration point
tc
Total runoff hydrograph (TRH)
Discharge Q
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.)
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 2 Storm 3
Storm 1
Intens
Pnet2 Pnet
Pnet1 3
ity i
3 peaks due to
3-storms in
succession
Recession
Discharge Q
lines
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,
11
HYDROGRAP – d St = Qt dt HYDROLO11
12
HYDROGRAP HYDROLO12
∴ St = – z Qt dt = – z (Q0 Kr–1) dt
∴ St =
Qt
log e K r log a z ax =
a x
...(5.2)
Hence, the discharge at any time is proportional to the water remaining in storage i.e.,
Q0 S0
= S ...(5.2 a)
Q
t t
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.
e
(cumec)
b
d g
f
h
i
b. Semi-log plot of
master depletion curve
Lower ends of curves
Log
Master depletion
curve (natural scale)
Discharge Q
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
Master depletion curve
14
Discharge Q (cumec 10 or
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)
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.
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.
3
Q0 179 × 86400 m /day
S0 = = = 2.75 × 10 8m
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 S0
=
Qt St
Qt
or St = S0 ×
Q
0
2 –t
Qt = Qo Kr
Log Qt = log Qo – t log Kr
1000
ý
ü
ý
ü
ý
ü
8 y c + x m (Slope m is –ve)
6 m = – log Kr, Kr = daily depletion factor
L or recession constant
4
Log-
3
Interflow
2
Surface
recession
10
8 H
6
5
4
3
F
1 M G
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
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)
12
HYDROGRAP HYDROLO12
Peak N days
B
I 0.2
N (days) = 0.84 A (empirical)
2
A = Drainage area (km )
Discharge Q
Direct
surface Four separation procedures
runoff
Point of F
rise E (N days after peak)
A 4 3 Recession
2 G
1 C
2 Ground water contribution
D to stream (base flow)
Time t (hr or days)
where A = area of the drainage basin, km 2 and the size of the areas of the drainage basin as a
guide for the values of N are given below :
(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.
12
HYDROGRAP HYDROLO12
Discharge
hydrograph
Surface
Discharge Q
runoff
Recession
curves E
Increase in base flow
A after flood
Bank storage
High water level Permeable
banks of stream
GWT
Ground water flow
Ground water flow
(during low water period)
Discharge
Direct hydrograph
Discharge Q
surface
runoff
Separation line
Recession
curve
Base flow Extrapolated
depletion curve
Time t (hr)
Rainfall, P
intensity i
Rainfall
Pnet = 26.6 cm = Runoff
c.g. of Pnet
f-index
Time t (min)
tr = 3 hr = Effective duration of storm
300
tp = 10 1 hr peak flow
N = 0.84 (40) = 1 3 day (too long)
0.2
2
Basin lag 4
SQd.t 986(3 ×60 ×60) m 3
250 Pnet = A = 6 2
40 × 10 m
Total run = 0.266 m
off or 26.6 cm ... 1
200 hydrograph
SQd = SDRO = 986 cumec
Discharge (Stream flow) Q
Pnet = P – Losses
150
Direct surface = S(i – )t ... 2
runoff = 26.6 cm
t = duration of rain (min)
for which i > f P from 1 & 2
100 120 = TRO = Q net
75 = DRO = TRO – BFO should balance
= Qd = 120 – 45 DSR stops recession
Point of
rise D.SR Recession
50 starts
net
r
t Pnet
6
Area under tr = 3 hr = unit hydrograph duration
4 curve = 1 cm
of runoff t
DRO Time to peak = t + r
2 2.82 = UGO = p
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
(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).
12
HYDROGRAP HYDROLO12
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)
Σ Qd 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.
12
HYDROGRAP HYDROLO12
Unit
period 1
i=
Rainfal
tr cm/hr tr
Pnet = 1 cm = i × tr
l
Recession time Tr
Lag time
tp
Qp tr/2 Peak of UG, Qp cumec
Time to peak
Concen-
tration time I, inflection
tc
point
Discharge Q
Unit
hydrograph
Area under
curve
º 1 cm of Channel
runoff over storage
catchment
area A
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.
i = t1 , xi = t x cm/hr
Intensi
1× t = 1 cm
r r r
x cm ü rt
ty
xi ýP x
i 1 cm þ net × tr = x cm
tr
tr
x Qp
Hydrograph (produces
Run off x cm of runoff)
= x cm Peaks occur
Qp at same time
Discharge Q
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 tr
Continuous storm split into two unit storms
Rainfal
y cm
x cm z cm
2
1 3
l
UGO × y
Discharge Q
1
2 3 Hydrographs
corresponding
to storm 1, 2 & 3
tr – hr UGO UGO × z
UGO × x
Time t (hr)
(b) Principle of superposition
(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).
13
HYDROGRAP HYDROLO13
y
y cm = i =
cm/hr S Mid–ordinates of unit periods
Pnet tr
intensity i
Rainfall
x = 0.5 + 2.0 + 4.7 + 7.1 + 5.6
x cm = i = cm/hr + 3.5 + 2.5 + 1.6 + 0.9 + 0.3
tr = 28.7 units
1
1 cm = i = cm/hr % of runoff in unit period 1
tr 0.5
Unit period = × 100 = 1.74 » 2 (rounded)
28.7
12
25% runoff and like that.
11 occurs
in unit 4
10 period 4 UGO × y = y cm runoff hydrograph
25%
Discharge Q (cumec) (say, 10 cumec
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 +
5 16% UGO = 1 cm runoff unit hydrograph
–
4
7.1 units 12% Area under UG = Area under bar graph
3 UGO Mid-ordinates of unit periods
5.6
4.7 (1 cm) 9%
2 7 3.5 Bernard’s distribution graph
5.5%
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)
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 km 2, 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.
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
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
1
TRO—Total runoff ordinate = gauged discharge of stream
2
BFO—Base flow ordiante read from graph separation line shown in Fig. 5.10 (a). N = 0.83 A0.2
= 0.89(40)0.2 = 1.73 days = 1.73 × 24 = 41.4 hr from peak, which is seen not applicable here; hence an
arbitrary separation line is sketched.
3
DRO—Direct runoff ordinate = TRO—B.F.O.
4
UGO—Unit hydrograph ordinate
DRO ΣDRO . t 986(3 × 60 × 60)m
3
=
Pnet ; Pnet = = 6 2
A 40 × 10 m
= 0.266 m = 26.6 cm
5
Time 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.
13
HYDROGRAP HYDROLO13
intensity i
Rainfall
Pnets due to 3-storms
y cm
x cm z cm
tr tr tr
Q3
Q2
Observed hydrograph
z
Q1 = x u1
Q2 = x u2 + y u1
Q3 = x u3 + y u2 + z u1, etc
Q4
y
Discharge Q
Successively solve
for u1, u2, u3, etc.
z which give tr – hr UGO
y
Q tr – hr
1 5
Q UGO z
x
Q6 x u3
UG
x
Q7 u2 z u4
y
u1 u3 us y u
u4 x u5
x
0 5
0 z
0 0
Time t (hr)
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
Table 5.2 Derivation of a 6-hour unit hydrograph from a complex storm, Example 5.2.
HYDROGRAP
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
0 u0 = 0 0 — — 0 = 20 – 20 u0 = 0
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
30
4u7 3u5
3u6
4u7 + 3u5 = 45.5 – 20 ∴ u5 = 8.5
3u6 = 29 – 20 ∴ u6 = 3
U| check for
UGO derived
V | above
33 3u7 3u7 = 20 – 20 ∴ u7 = 0
W
*Except the first column, all other columns are in cumec.
**x = 2 cm, y = 4 cm, z = 3 cm
1
13
HYDROGRAP HYDROLO13
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)
∴ 6 = 0.01 m, or 1 cm
118.8 × 10
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.
4 cm 3 cm
2 cm
yi
6 hr 6 hr 6 hr
200 220
180 200
160 a. observed
180
3
hydrograph due
140 to 18-hr storm
160
(u’s = 6-hr UGO)
3
120 140
Surface runoff Q
Stream flow Q
100 120
4
80 100
UGO × 4
3
60 80
4
40 60
2
UGO × 3
3
UGO × 2
u3
2
20 40
u5
4
2
3
2
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)
13
HYDROGRAP HYDROLO13
40
35
30
6-hr UG derived
25
(u's = UGO)
Discharge Q
20
u
15
10
u
5
u
u
0
0 3 6 9 12 15 18 21
Time t (hr)
Lp 1 0 0 0 0 0 O Lu O 1 Lq O 1
M pp 2 p1 0 0 0 0
P Muu P 2
Mqq P 2
U=M P Q=M P
3 p2 p1 0 0 0 3 3
where P =
M P
0 p3 p2 p1 0 0
Mu P
4 Mq P
4
MN0M0 00
p3
0
p2
p3 00 0
0 P MN0M0 P q
MNq PQ
5
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
P T 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
13
HYDROGRAP HYDROLO13
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
t1
tpeak
t2
t3
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.
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)
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
13
HYDROGRAP HYDROLO13
20
18
16 Combined hydrograph
(2 cm in 6 hr)
14
3-hr UG's lagged by 3 hr
12
Discharge Q
10
6-hr UG (1 cm in 6 hr)
(comb. hyd. ord. ¸ 2)
8
2
0
0 3 6 9 12 15 18 21 24 27
Time t (hr)
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.
Intensity
1 cm 1 cm 1 cm 1 cm 1 cm 1 cm To obtain tr – hr UG
tr tr tr tr tr tr multiply the S-curve
480
difference by tr/tr ¢
440
400 S-curve A
360 hydrograph
320 Lagged
S-curve
280 (lagged by Area
s-curve Constant outflow 2
tr – hr) difference
240 Unit hydrographs in 2.78 A(km )
Discharge Q
Qe =
= 1´ t cm succession produce tr (hr)
200 t r
r
in tr – hr constant outflow, Qe cumec 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)
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.
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
1
Time 4-hr UGO S-curve additions S-curve lagged S-curve 2-hr UGO
(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
0 0 — — — — — 0 — 0 0
2 25 — — — — — 25 0 25 50
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*
HYDROLO
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
Fig. 5.19.
14
HYDROGRAP HYDROLO14
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.
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
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
480
440
400
0
0 2 4 6 8 10 12 14 16 18 20 22 24 26
t r¢
Time t (hr)
= 2 hr lag
Fig. 5.19 Derivation of 2-hr & 6-hr UG from a 4-hr UG (Example 5.4 & 5.5)
Total stream flow at concentration point on successive days are 19.6, 62.4, 151.3, 133.0,
89.5, 63.1, 43.5, 28.6, and 19.6 cumec.
Estimated base flow during the corresponding period on successive days are 19.6, 22.4,
25.3, 28.0, 28.0, 27.5, 25.6, 22.5 and 19.6 cumec.
14
HYDROGRAP HYDROLO14
1 2 3 4
18
1 18 4.5 = × 100
402
2 96 24
3 120 30
4 82 20
5 47 12
6 25 6
7 12 3
8 2 0.5
Pnet = x cm
ity i
Pnet =1cm
tr
Unit period
25% Distribution % = 0i × 100
S0i
19.5%
– +ve & –ve areas should balance
16 % (area under bar graph = area under UG)
tr – hr UG
02 07 3%
2% 08
01 09 0 1%
10
1 2 3 4 5 6 7 8 9 10
Time base T divided into 10 unit periods
Unit periods
Solution The total runoff hydrograph and estimated base flow are drawn in Fig. 5.21 and
the direct runoff ordiantes on successive mid-days are determined as DRO = TRO – BFO and
the percentages of direct runoff on successive days computed in Table 5.6. Column (3) gives the
distri-
bution percentages, and the derived distribution graph for 1-day unit storms is shown in Fig.
5.22.
160
150
140
130
120
100
80 82
70
60 Direct surface
runoff
50 47
40 25
30 18 12
2
20
Base flow
10
0
0 1 2 3 4 5 6 7 8 9 10 11
Time t (days)
20 Distribution
20 percentages
16
12
12
8
6
4.5
4 3
0.5
0 1 2 3 4 5 6 7 8
1-day unit periods
1000
Runoff hydrograph
due to 2-storms on
600 July 15 & 18
Direct surface runoff
400
200
0
0 1 2 3 4 5 6 7 8 9 10 11 12
July 15 16 17 18 19 20 21 22 23 24 25 26
Days (date)
1
Unit periods (day)
1 2 3 4 5 6 7 8 9 10 11
Total rainfall (cm) → 7 10
Loss of rain (cm) → 2 2
HYDROLO
HYDROGRAP 14
The base width is 12 hr or 6 unit periods. As a first trial, try a set of six distribution
percentages of 10, 20, 40, 15, 10, 5 which total 100%. The direct surface discharge can be
converted into cm/hr as
6
1 3210 × 10
1 cm/hr = × = 8920 cumec
100 60 × 60
5 cm/hr
Intens
T
T
T
TDSR
Time t (hr)
2.5 Gauged
Direct surface runoff
1.5
1.0
0.5
0
0 1 2 3 4 5 6 7 8
1 10, 5 2-hr-unit periods
20, 20
Distribution
2
3 40, 40
4 15, 20
%
5
10, 10
6
5,
Fig. 5.25 Derivation of distribution percentages (Example 5.7)
Table 5.8 Derivation of distribution percentages by trials (Example 5.7)
1
2-hr-unit periods
1 2 3 4 5 6 7 8
Rainfall rate (cm/hr) → 2.5 6.5 4.5
Loss rate (cm/hr) → 1.5 1.5 1.5
Net rain (cm/hr) → 1 5 3 (= 9 × 2 cm)
Unit distribution Distribution (cm/hr)
Periods Percentage
1 10 (5) (0.05) (0.25) (0.15)
0.10 0.50 0.30
2 20 0.20 1.00 0.60
3 40 0.40 2.00 1.20
4 15 (20) (0.20) (1.00) (0.60)
0.15 0.75 0.45
5 10 0.10 0.50 0.30
6 5 0.05 0.25 0.15
Total 100 (100) cm/hr 0.10 0.70 1.70 2.75 2.05 1.00 0.55 0.15 (= 9 × 2 cm check)
(0.05) (0.45) (1.55) (2.80) (2.30) (1.15)
cumec 446 4015 1382 25000 20520 10260 4900 1338 (446/8920 = 0.05 cm/hr)
Note Figures in brackets indicate the adjusted values in the second trial.
HYDROLO