Runoff

Runoff
Draining or flowing off of precipitation from a catchment area
through a surface channel

The output from the catchment in a given unit of time

• For a given precipitation,

– evapotranspiration,
– initial loss,
– infiltration and
– detention-storage requirements

have to be first satisfied before the commencement of

Runoff
Runoff
• Overland flow - excess precipitation
moves over the land surfaces to reach
smaller channels

• Surface runoff - Flows from several

small channels join bigger channels and
flows from these in turn combine to form a
larger stream, and so on, reaches the
catchment outlet
 Interflow/seepage/subsurface/storm
flow or Quick return flow

A part of the precipitation that infiltered moves laterally

through upper crusts of the soil and returns to the
surface at some location away from the point of entry
into the soil
– Depends on geological conditions of the
catchment (physiographic feature)

• A fairly pervious soil overlying a hard impermeable surface

is conducive to large interflows

• Depending upon time delay between the infiltration and the

outflow, the interflow is classified
– Prompt Interflow
– Delayed interflow
 Groundwater Runoff/flow
• Groundwater flow- another route for the infiltered water is to undergo
deep percolation and reach the groundwater storage in the soil

• The groundwater follows

a complicated and long path of travel, and ultimately reaches the surface
(with large time lag, being of the order of months and years)
provides dry-weather flow in perennial streams
 Classification of Runoff
• Based on the time delay between the precipitation
and the runoff

– Direct Runoff: part of runoff which enters the stream

immediately after the precipitation
• Surface runoff
• Prompt interflow
• Precipitation on the channel surface

– Base Flow: The delayed flow that reaches a stream

essentially as groundwater flow
 Streamflow
• A stream can be defined as a flow channel into which
surface runoff from a specified basin drains.

• Stream-flow represent the runoff phase of the hydrologic

cycle.

• Stream-flow is amenable to fairly accurate measurement

in contrast to evaporation, precipitation,
evapotranspiration etc.

• There is considerable exchange of water between a

stream and underground water.
 Stream-flow measurement
techniques (Hydrometry)
• Direct measurement
– Area-velocity method
– Dilution technique
– Electromagnetic method
– Ultrasonic method

• Indirect measurement
– Hydraulic structures, such as weirs, flumes and gated
structures
– Slope area method
 General
• Stream-flow is measured in units of discharge (m3/s) at a
specified time and constitute historical data

• Continuous measurement of discharge is very difficult.

• Direct measurement of discharge is very time consuming and

costly.

• Two step procedure is followed

– Discharge in a given stream is related to elevation of the
water surface (stage) through a series of careful
measurements.
– Stage of the stream is observed routinely and discharge is
estimated by stage discharge relationship.

• Have you seen a scale in river bridge piers???

Stage discharge relationship

stage discharge relationship curve

 Measurement of stage
• It is defined as the water surface elevation
measured above a datum (can be MSL or any
arbitrary datum).

• Manual stage measurement

– Staff gauge
– Wire gauge

• Automatic stage measurement

– Float gauge recorder
– Bubble gauge
 Staff gauges

Vertical Staff Gauge Sectional Staff Gauge

Float gauge recorder

Float Gauge recorders

 Stage data presentation
• Presented in the form of plot
of stage against
chronological time, known
as stage hydrograph

• Uses
– Determination of stream
discharge
– Flood warning and flood
protection works
– Reliable long term data to
peak floods can be analyzed
statistically to estimate design
peak river stages for use in the
design of the hydraulic
structures, such as bridges,
weirs etc.
 Stage Discharge relation ship (rating
curve)
• First a stage discharge relationship has to be
developed to use the stage data

• Thus the main aim of measuring stream-flow is to

establish this rating curve

• Let G be the stage and Q be the discharge. Once the

relationship is established, later just measure only the
stage and read the discharge from rating curve.

• First plot the discharges against the stages measured.

– If the relationship is permanent and does not vary with respect
to time it is called permanent control else it is called shifting
control
Stage discharge curve
 Relationship (for Permanent Control)

Q=Cr(G-a)
Where Q –discharge, G - is the stage, a – is a constant which represent
the gauge reading corresponding to zero discharge (normally
estimated by extrapolating in the plot, since the measurement is
almost very difficult)
Cr and  are rating curve constants

The best value are estimated either by method of least square or by

graphical method

logQ= log(G-a)+logCr
Plot the measured values and find the constants
The above equation is called as rating equation.
Step 1
Solve? Develop a stage discharge relationship for the
measured data. Find stage for a discharge of 2600
m3/s.
Stage Discharge Step 1
(m) (m3/s) ln(G-a) ln Q 26.00

21.95 100 -0.654 4.606

22.45 220 0.020 5.394 25.00

22.8 295 0.315 5.688

23 400 0.451 5.992 24.00

Stage (m)
23.4 490 0.678 6.195
23.57 500 0.761 6.215
23.00
23.65 640 0.798 6.462
24.05 780 0.963 6.660
22.00
24.55 1010 1.138 6.918
24.85 1220 1.230 7.107
21.00
25.4 1300 1.379 7.171
0 200 400 600 800 1000 1200 1400 1600 1800
25.15 1420 1.314 7.259 Discharge (cu.m/s)

25.55 1550 1.416 7.347 But actual measured is 20.5 m

25.9 1760 1.498 7.474 only????????
From the graph the value of a (discharge for zero stag is 21.50 m)
 Contd…
1.50

(G-a) =
Fit Results

Fit 1: Linear, Y=B*X+A

Equation:
1.00 Y = 0.726826 * X + -3.89026

0.0205Q0.7268
Number of data points used = 14
Average X = 6.46345
Average Y = 0.807547
Regression sum of squares = 4.83243
Residual sum of squares = 0.0538442
Coef of determination, R-squared = 0.988981
0.50 Residual mean square, sigma-hat-sq'd = 0.00448702

Stage (m)
G=27.63 m for a discharge
of 2600 m3/s
0.00

(G-a) =
0.163Q0.466 -0.50

G=26.85 m for measured a

of 20.5 m -1.00

Else??
0 1 2 3 4 5 6 7 8 9 10
Discharge (cu.m/s)
 Contd…
• This method holds good if the flow is virgin flow

• Stream-flow should be purely from runoff only

• If there is reservoir release or

• There is a back water effect

• Then we need site specific relationship and the

available rating curve need to be either
extrapolated or modified accordingly
First step in measuring the discharge is
the Measurement of velocity
• Commonly measured by a mechanical device called current meter.

• It measures the velocity at a point in the flow.

• It consists of a rotating element which rotates due to reaction of the

stream current with an angular velocity proportional to the stream
velocity.

• Two main types of current meter

– Vertical axis meters
– Horizontal axis meters
Vertical axis current meter
• Approximate stream velocity
can be determined by floats.

Horizontal axis current meter

 Current meter
• It is so designed that its rotation speed varies linearly
with the stream velocity at the location of the
instrument.
v = a Ns + b
Where
v = stream velocity at the instrument location in m/s
Ns = revolutions per second of the meter
a, b = constants of meter

• Each instrument has a threshold velocity below which

above equation is not valid.

• No of revolutions are counted for a known interval of

time.
 Horizontal axis meters

Vertical axis meters

 Calibration of Current meter
• The relation between stream velocity and
revolutions per seconds of the current meter is
called calibration equation.

• Calibration equation is unique to each

instrument.

• It is determined by towing the instrument in a special tank.

• The towing tank is a long channel containing still water with

arrangements for moving a carriage longitudinally over its surface at
constant speed.

• The instrument to be calibrated is immersed to a specified depth.

• Average value of revolutions per seconds (Ns) are determined for a

predetermined constant speed (v). This is repeated over a complete
range of velocities and a best fit linear relationship is formed.
 Velocity measurement by Floats
• Float method - the basic idea is to measure the time that it takes for
an object to float a specified distance downstream.
• It is inexpensive and simple.
• This method measures surface velocity.
S
vS 
t
where vs = surface velocity, S = distance traveled in time t

• Mean velocity is obtained using a correction factor.

vmean = k vs
where k is a coefficient that generally ranges from 0.8 for rough beds to
0.9 for smooth beds (0.85 is a commonly used value)

• Useful for
– A small stream in floods
– Small stream with rapidly changing water surface
– Preliminary surveys
 Area velocity method
• It consist of the measurement of the area of cross
section of the river at a selected section and measuring
the velocity through it.
Q = AV

• Criteria for gauging site

– The stream should have a well-defined cross-section which does
not change in various seasons.
– It should be easily accessible all through the year.
– The site should be in a straight, stable reach.
– The gauging site should be free from backwater effects in the
channel.
 Contd…
• Cross section area is divided into a large no of cross subsections by
verticals.

• The velocity in these subsections are measured by current meter.

• Accuracy of discharge measurement increase with no of

subsections which also increase time, effort and expenditure .

• Guidelines for selection of subsections

– The segment width should not be greater than 1/15 to 1/20 of the width
of the river.
– The discharge in each segment should be less than 10% of the total
discharge.
– The difference of velocities in adjacent segments should not be more
than 20%.
Cross-section of a stream to measure the
discharge
Variation of stream velocity with
depth

Variation of stream velocity with depth

 Steps in finding the discharge after
measuring the area and velocity
• If qi is discharge in elemental area
• Width is given by (Wi+Wi+1)/2
• If di is depth at velocity measurement
• Area of element is ai= di* (Wi+Wi+1)/2
» for I = 2, 3, 4, …..n-2
• For segment 1 and N it is triangle
 2   2 
 W  W2     W  W N 1  
 1 2    N 2  
a1  d1   aN  d N   
2W1  2W N 
 
   
   
• Discharge in elemental area qi=ai*vi
• Total discharge is N
Q   a i vi
i 1
 Solve?
The data for a stream gauging station is as follows. The rating equation
of current meter is given as v  0.51N s  0.03 m / s
Ns is number of revolutions per second
Distance from left Depth (m) Number of revolution of a Duration of
edge (m) current meter kept at observation
Average Velocity Segmental
0.6 m depth
width (m) (m/s) discharge cu.m/s

0 0 0 0 0 - -
1 1.1 39 100 2 0.229 0.504
3 2 58 100 2 0.326 1.303

5 2.5 112 150 2 0.411 2.054

7 2 90 150 2 0.336 1.344

9 1.7 45 100 2 0.260 0.882

11 1 30 100 2 0.183 0.366

12 0 0 0 0 - -

6.453
 Moving-Boat Method
• In large rivers standard
current meter is very time
consuming and it is difficult to
keep the boat stationary.

• Propeller type current meter is

used which is free to move
about its vertical axis is used.

• If Boat velocity is Vb, Flow

velocity is Vf and Resultant
velocity is VR (meter align
itself in the direction of
resultant velocity)
vb = vR cos θ and vf =vR sin θ
Moving boat method
 Contd….
• If the time of transit between two verticals is Δ t, then the width
between the two verticals
W = vb Δ t

• The flow in the sub-area between two verticals i and i+1 where the
depths are yi and yi+1 respectively, by assuming the current meter
to measure the average velocity in the vertical, is

 Stream discharge

Q = Σ ΔQi
 Solve
? The following data has been obtained in a moving boat.
Determine the discharge in the stream. The sections are
spaced at a constant distance of 75 m apart.
Angle made by current meter
Resultant velocity with the direction of boat
Section (m/s) movement (degrees) Depth (m)
0 Right bank - 0
1 1.75 55 1.8
2 1.84 57 2.5
3 2.00 60 3.5
4 2.28 64 3.8
5 2.30 65 4.0
6 2.20 63 3.8
7 2.00 60 3.0
8 1.84 57 2.5
9 1.70 54 2.0
10 Left bank - 0.0

Discharge = 3458 cu.m/s

Dilution technique for stream flow
measurement
• It depends upon the continuity principle applied to a
tracer which is allowed to mix completely with the flow.
– For a constant injection method if
Co = small initial concentration of the tracer in the stream-flow
C1 = high tracer concentration at section 1 where tracer is added
C2 = tracer concentration at section 2 where tracer mixes
thoroughly with the fluid due to the turbulent mixing process
Qt=discharge of tracer and Q-is discharge in the stream

i.e.
 Solve
?A 500 g/l tracer was injected at a constant rate of
4l/s. At downstream length the concentration of
tracer was measured as 4 ppm. If initial tracer
concentration in the stream was zero. Estimate
the stream discharge?
1 gram/liter [g/L] = 1001.142303 part/million [ppm]
The parts per million (ppm, 10 ⁶) is a pseudo unit that describes small
values of miscellaneous quantities, for example, mole fraction or
mass fraction. These fractions are dimensionless because they are
quantity-per-quantity measures. Therefore, they are simply
numbers. One part per million denotes one part per 1,000,000 parts
and a value of 1 × 10 ⁶. This can be compared to one drop of a
liquid substance diluted into 50 litres of water. Other part-per units
are commonly used: ppb (parts-per-billion, 10 ⁹), ppt (parts-per-
trillion, 10 ¹²) and ppq (parts-per-quadrillion, 10 ¹⁵).
 Discharge Q = 500 m3/s
 Tracer
• Ideal properties of tracer:
– It should not be absorbed by the sediment, channel boundary
and vegetation. It should not chemically react with any of the
above surfaces and also should not be lost by evaporation.
– It should be non-toxic.
– It should be capable of being detected in a distinctive manner
in small concentrations.
– It should not be very expensive.

• Types of tracer:
– Chemicals (common salt and sodium dichromate )
– Fluorescent dyes (Rhodamine-WT and Sulpho-Rhodamine B
Extra )
– Radioactive materials (such as Bromine-82, Sodium-24 and
Iodine-132) (radioactive tracers).
Runoff and Streamflows
- Hydrographs
- Rainfall-Runoff Models
 Hydrograph
A plot of the discharge in a stream plotted against
time chronologically

10
11
12
13
14

24
15
16
17
18
19
20
21
22
23
25
26
27

34
28
29
30
31
32
33
35
36
37
38
39
40
42
41

45
43
44
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67

70

73
68
69
71
72
74
75
76
77
78
79
80
1
2
3
4
5
6
7
8
9
0 12

Rainfall
5 Crust 10

Time Lag 10
Recession 8
Limb
Runoff (V/t)

15

20

Peak Discharge 4

25

Base flow line

Rising 2
30
Limb
35 0

Time
 Components of Runoff
• Based on the source from which runoff is derived, the total runoff is visualised
to consist of surface runoff, subsurface runoff and groundwater runoff

42
• Depending upon the unit of time involved

– Annual hydrographs: daily or weekly or 10 daily

mean flows over a year

– Monthly hydrographs: daily mean flows over a

month

– Seasonal hydrographs: discharge in a particular

season such, as the monsoon season or dry season;
and

– Flood hydrographs or hydrographs due to a storm

representing stream flow due to a storm over a
catchment
• Each of these types have particular
applications

– Annual and seasonal hydrographs

• Calculating the surface water potential of stream
• Reservoir studies
• Drought studies

– Flood hydrographs
• Analyzing stream characteristics associated with
floods
 Water Year
• In annual runoff studies it is advantageous to
consider a water year

– Time when the precipitation exceeds the average

evapotranspiration losses

– In India, June Ist - May 31st of the following calendar year

– Complete cycle of climatic changes is expected and

hence the water budget will have the least amount of
carry over
 Runoff Characteristics of
Streams
• Classification of stream based on
annual hydrographs of streams

– Perennial

– Intermittent

– Ephemeral
 Perennial Stream
which always carries some flow
• There is considerable amount of groundwater flow throughout the year
• Even during dry seasons the water table will be above the bed of the stream
 Intermittent Stream
• During the wet season the water table is above the stream bed and
there is a contribution of the base flow to the stream flow

• During dry seasons the water table drops to a level lower than that of the
stream bed and the stream dries up
 Ephemeral Stream
• Does not have any base-flow contribution

• The annual hydrograph, shows series of short-duration spikes marking flash

flows in response to storms

• Becomes dry soon after the end of the storm flow

Characteristics of Streams based on
ground water contribution
– Effluent streams

– Influent streams
Flow characteristics of a stream
depend upon
• Rainfall characteristics
– Magnitude intensity
– Distribution time
– Space and its variability

• Catchment characteristics
– Soil
– Vegetation
– slope, geology, shape
– Drainage density

• Climatic factors which influence

evapotranspiration.
 Factors Affecting Runoff
Climatic Factors Physiographic factors
Type of precipitation Land use
Intensity of rainfall Soil Type
Duration of rainfall Area of basin
Areal distribution of rainfall Shape of basin
Direction of storm movement Elevation
Antecedent precipitation Slope
Other climatic factors that affect Orientation
evaporation and transpiration
Type of drainage network
Indirect Drainage
Artificial drainage

52
 YIELD
(ANNUAL RUNOFF VOLUME)
The total quantity of water that can be expected from a
stream in a given period
• The calculation of yield is of fundamental importance in all
water-resources development studies
• The water yield of a watershed is the amount of freshwater
generated from a combination of base flow, interflow and overland
flow originating from groundwater, precipitation and/or snowpack
• Methods to estimate yield
– Empirical equations
– Correlation of stream flow and rainfall,
• Time series modeling, ANN, GP and MT models
– Watershed simulations
Block Box Models
Conceptual models
Physical models
 Dependable Yield
• The annual yield from a basin is the effect of various processes, which are
random in nature. Therefore the annual yield is also a random variable
whose magnitude varies from year to year.

• The annual yield in any year in future can be estimated only with some
degree of dependability associated with it.
– Thus for example, 75% dependable yield of a basin is the annual runoff
volume from the basin which can be expected to be equalled or exceeded,
on the average, 3 years out of 4 years.
– Similarly 50% dependable yield can be expected, on the average, every
alternate year.

• If the annual yield of a basin is measured for a sufficiently long period of

time, the yield corresponding to any stipulated level of dependability can be
determined using the frequency analysis or from an appropriate probability
distribution fitted to the observed data.
 Example
• The annual yields of a basin in million m3 observed for a
period of 19 years from 1971 to 1989 are 832, 672, 488,
1632, 608, 1128, 960, 560, 880, 984, 640, 832, 696, 632,
896, 464, 688, 432 and 1440 respectively. Using the
frequency analysis determine the following.
(a) 75% dependable yield
(b) 50% dependable yield
(c) Dependability of yield of 1300 million m3.

55
Solution

56
 Continued…

• From the above figure :

i. 75 % dependable yield = 536 Mm3
ii. 50 % dependable yield = 720 Mm3
57
iii. Dependability of yield of 1300 Mm3 = 12%
 Runoff/Yield Models

• Methods to estimate yield

– Empirical equations
– Regression equations - Correlation of stream flow & rainfall
• Linear/nonlinear regression, ANN, GP and MT models
– Watershed simulations
Black Box Models
Conceptual models
Physical models
 Types of Rainfall-Runoff models Models

• Empirical models: based on observational data and

calibrated input-output relationship without description of
individual processes

• Conceptual models: basic processes (evaporation, etc.) are

separated to some extent, but their algorithms are
essentially calibrated input-output relationships

• Physically based models: based on reliable relationships

between watershed characteristics and the parameters
 Empirical Equations

• Binnie's Percentages
• Barlow's Tables
• Strange's Tables
• Inglis and DeSouza Formula
• Khosla's Formula

SCS-CN technique
 Binnie's Percentages
• Sir Alexander Binnie measured the runoff from a
small catchment near Nagpur (Area of 16 km2 )
during 1869 and 1872

• Developed curves of cumulative runoff against

cumulative rainfall

• Established percentages of runoff from rainfall

• These percentages are beening used in Madhya

Pradesh and Vidarbha region of Maharashtra for
the estimation of yield
 Barlow's Tables
• Barlow, the first Chief Engineer of the Hydro-
Electric Survey of India (1915)

• Conducted study on small catchments (area~

130Km²)in Uttar Pradesh expressed runoff R as
R = Kb P

where Kb = runoff coefficient

– which depends upon
• type of catchment
• nature of monsoon rainfall.
 Barlow’s Runoff Coefficient Kb in Percentage

Class Description of catchment Values of Kb (percentage)

Season Season Season
1 2 3
A Flat, cultivated and absorbent soils 7 10 15
B Flat, partly cultivated and stiff soils 12 15 18
C Average catchment 16 20 32
D Hills and plains with little cultivation 28 35 60
E Very hilly, steep and hardly any 36 45 81
cultivation
Season 1: light rain, no heavy downpour
Season 2: Average or varying rainfall, no continuous downpour
Season 3: Continuous downpour
 Strange's (1928) Tables
• Data on rainfall and runoff in the border areas, of Maharashtra and
Karnataka and obtained the values of the runoff coefficient

Ks = R/P
as a function of the catchment character

• Catchments were characterized as "good", "average" and "bad'.

• Strange also gave a table for calculating the daily runoff from daily
rainfall.

• In this the runoff coefficient depends not only on the amount of

rainfall but also on the state of the ground.

• Three categories of the original ground state as 'dry', 'damp' and

'wet' are used by him
Extract of Strange’s Table of Runoff Coefficient Ks in
Percent

Total monsoon Runoff coefficient Ks percent

rainfall (cm)
Good Catchment Average Bad Catchment
Catchment
25 4.3 3.2 2.1

50 15 11.3 7.5

75 26.3 19.7 13.1

100 37.5 28 18.7

125 47.6 35.7 23.8

150 58.9 44.1 29.4

 Inglis and DeSouza (1929)
Formula
• Stream gauging in 53 sites in Western
India resulted, two regional formulae
between annual runoff R in cm and annual
rainfall p in cm as follows:
• For Ghat regions of western India
R = 0.85 P - 30.5
• For Deccan plateau
1
R PP  17.8
254
 Khosla's Formula
• Monthly data on rainfall, runoff and temperature data for
various catchments in India and USA considered

Rm = Pm - Lm
Lm = 0.48 Tm for Tm > 4.5°C

where Rm = Monthly runoff in cm and Rm ≥ 0

Pm = monthly rainfall in cm
Lm = monthly losses in cm
Tm = mean monthly temperature of the catchment in °C

• For Tm ≤ 4.5°C, the loss Lm may provisionally be assumed as

ToC 4.5 -1 -6.5
Lm (cm) 2.17 1.78 1.52

Annual runoff = Rm

 Disadvantages of Conceptual and
Physically based Models
• Data Intensive
• Complex and large number of parameters are
involved
• Implementation and calibration of such a model
can typically present various difficulties
• Requiring sophisticated mathematical tools,
significant amounts of calibration data and
some degree of expertise and experience with
the model.
• Data on the physical characteristics of the
catchment are available at the grid scale is
required
Runoff and Streamflows
- Rainfall-Runoff Models
- Hydrological models
 Watershed Simulation
• Water-budget equation for the determination of
runoff
Rs = P - Eet – f - Δ S
R = Rs + Go
Rs = surface runoff,
P = precipitation,
Eet = actual evapotranspiration,
G0 = net groundwater outflow and
ΔS = change in the soil moisture storage
f = infiltration loss

The sum of Rs and Go is considered to be given by

the total runoff R, i.e. stream-flow
Runoff and Streamflows
- Rainfall-Runoff Models
- Hydrological models
 Watershed Simulation
• Water-budget equation for the determination of runoff
Rs = P - Eet – f - Δ S
R = Rs + Go
where Rs = surface runoff,
P = precipitation,
Eet = actual evapotranspiration,
G0 = net groundwater outflow and
ΔS = change in the soil moisture storage
f = infiltration loss

The sum of Rs and Go is considered to be given by the total runoff R

g i.e. stream-flow
 Rainfall-Runoff Modeling
Data required for RR Modeling
• The data required to develop a model for particular watershed
includes
– Soil characteristics,
– initial soil moisture,
– infiltration, and
– groundwater flow

• But these data are not easily available as compared to the rainfall (P)
and runoff (Q) data.
• In most of the studies, variable ‘P’ is selected in order to forecast Q.
• All the rainfall data are lumped with respect to time and a single time
series rainfall data will be used to predict the runoff
 Rainfall-Runoff (RR) models

Two simple RR models, but widely used RR models

1. Rainfall-Runoff Correlation
2. SCS-CN Techniques
 Two simple RR models, but widely
used RR models
1. Rainfall-Runoff Correlation n PR   P  R
a
R= aP+b
 2

n  P   P  2

R = runoff
P = precipitation
 R  a P
a = is the slope b
b = is the intercept
n
r = coefficient of correlation

n  PR   P  R
r
n P   P  X n R   R  
2 2 2 2
 Develop a linear RR model
Month P R Month P (cm) R
(cm) (m3/s) (m3/s)
1 5 0.5 10 30 8.0
2 35 10.0 11 10 2.3
3 40 13.8 12 8 1.6
4 30 8.2 13 2 0.0
5 15 3.1 14 22 6.5
6 10 3.2 15 30 9.4
7 5 0.1 16 25 7.6
8 31 12.0 17 8 1.5
9 36 16.0 18 6 0.5
 20
Fit Results

18 Fit 2: Linear, Y=B*X+A

Equation:
Y = 0.380197 * X + -1.55603
Number of data points used = 18
16 Average X = 19.3333
Average Y = 5.79444
Regression sum of squares = 405.607
14 Residual sum of squares = 30.5424
Coef of determination, R-squared = 0.929973
Residual mean square, sigma-hat-sq'd = 1.9089
Coef. of correlation, R = 0.964
Runoff (cu.m/s)

12

10

0 6 12 18 24 30 36 42
Rainfall in cm
Developing Multi Linear Regression Models
 In the linear regression model, the dependent variable is assumed to
be a linear function of one or more independent variables plus an
error introduced to account for all other factors.
 The multiple linear regression models are represented in equation
as:
 y = m1x1 + m2x2 + m3x3 +………+ mnxn + c
 The general multiple linear regression model considered is given in
equation
 Various models have been generated, by varying the value of m and
n in the equation and 26 different models have been developed.70
% of data used for training and 30 % used for testing
 Q (t+1) = f {P (t), P (t-1),…… P (t -m), Q(t)Q (t-1),……, Q (t - n)}
 2. SCS-CN Technique
• Developed by Soil Conservation Service of USA for
small agricultural watersheds, latter extended to urban
watershed.

• Data needed is the rainfall and type of soil and land

use

• Based on the concept of reservoir operation

– Let the water storing capacity of the soil (S)
• S-maximum retention of water in the soil – capacity of the reservoir
– Runoff will start only when this reservoir is full
But actual soil moisture is Sa = P-Ia-R

Similarly let
R (actual runoff)
P-Ia = Potential runoff
R =0 if P<Ia

The hypothesis of SCS-CN technique is that the ratio of actual and potential
moisture retained is equal to the ratio of actual and potential runoff

P  Ia  R R

S P  Ia

R
P  I a 2
By rearranging this the SCS-CN equation is
P  Ia  S

SCS-CN found that Ia = 0.2 S

R
 P  0.2 S 2
P  0.8S
• A number of curves has been
plotted for the measured P and
R

• SCS found a number of

dimensionless curves with
numbers called Curve
Numbers 0CN100

– For impervious and water

bodies CN=100
– It varies with respect to
antecedent soil moisture, soil
type, and land use.
– This CN is related to S
1000
S  10
CN
In fps system
P and R in inches
25400
S  254 For various land uses the CN is taken as
CN weighted curve Number
Where P and R in mm
 Runoff curve numbers for selected agricultural, suburban and urban land uses (antecedent moisture condition II, Ia = 0.2S)
Hydrologic Soil Group
Land Use Description
A B C D
without conservation treatment 72 81 82 91
Cultivated Land:
with conservation treatment 62 71 78 81
Poor condition 68 79 86 89
Pasture or range land:
Good condition 39 61 74 80
Meadow: Good condition
30 58 71 78
thin stand, poor cover, no mulch 45 66 77 83
Wood or Forest land:
Good condition 25 55 70 77
Good condition: grass cover on 75% or more of the
Open spaces, lawns, parks, golf area 39 61 74 80
course, cemeteries, etc.
fair condition: grass cover on 50% to 75% of the area
49 69 79 84
Commercial and business areas (85% impervious)
89 92 94 95
Industrial districts (72% impervious)
81 88 91 93
Average lot size Average % impervious
1/8 acre or less 65 77 85 90 92
1/4 acre 38 61 75 83 87
Residential
1/3 acre 30 57 72 81 86
1/2 acre 25 54 70 80 85
1 acre 20 51 68 79 84
Paved parking lots, roofs, driveways, etc. 98 98 98 98
Paved with curbs and storm sewers 98 98 98 98
Street and roads gravel 76 85 89 91
dirt 72 82 87 89
Soil Group Description of the Hydrologic Soil group by USSCS

A Lowest runoff potential. Includes deep sand with very little clay and silt.
B Moderately low runoff potential. Mostly sandy soil less deeper than A
C Moderately high runoff potential. Comprises shallow soil of clay, colloids.
D Highest runoff potential. Includes mostly clay of high swelling percent, nearly
impermeable soils.

Antecedent moisture 5-day antecedent rainfall (cm)

condition class Non-monsoon period Monsoon period
I. Optimum soil condition Less than 1.25 Less than 3.5
from plastic limit to wilting
point
II Average value for 1.25 to 2.75 3.5 to 5.25
annual floods
III. Heavy rainfall or low Over 2.75 Over 5.25
rainfall and low
temperature during five
days preceding the storm
 ? Determine the runoff from a watershed of 60 ha.
The following data is available
• Rainfall = 100 mm
• Row crop without conservation 40 ha (soil type C)
• Wood land good condition 20 ha (soil type B)

From the CN table find the CN corresponding to these two soil type and
land use
– For row crop 40 ha = 82
– For wood land 20 ha = 55
– The weighted CN = CN  (82 * 40)  (55 * 20) = 73
60

25400
S= S  254 = 93.9 mm
73

R
 P  0.2 S 2
= 37.7 mm
P  0.8S
 Hydrological Models

Hydrological Models

Empirical models Physical models

Lumped Models Distributed Models Semi-distributed Models

Fig. Classification of Hydrological Models

• Based on the representation of physical process hydrological models

can be classified as lumped, distributed and semi-distributed models
 Lumped Models
• A lumped model is a model that represents watershed as a single
homogeneous unit.
• Lumped models assume uniform soil types, vegetation types, and land use
practices over a watershed (Paudel, 2010).
• Key limitation of simple lumped models is that their parameters are not
directly related to the physical characteristics of the catchment.
– The applicability is limited to gauged watersheds where the expected
conditions are within the historical data used for calibration and where no
significant change in catchment conditions has occurred.
• Some of the lumped models are:
– Integrated Hydro Meteorological Model (IHMM) (Georyakakos et al, 1999),
– Illinois Urban Catchment Runoff Simulation (ILUCAT) (Ben et al., 1999),
– Geomorphology-Based Hydrological Model (GBHM) (Yang et al., 2007),
– Hydrologic Simulation Model (HSIM) (Chiew et al., 2007),
– Large Basin Runoff Model (LBRM) (Hunter et al., 2007),
 Semi Distributed Models
• The spatial variation of watershed characteristics can be considered partially in
semi distributed models.
• In this type of hydrologic modeling, the watershed is divided into smaller
computational elements or sub-basins, and hydrologic computation is carried out
for each element.
• The subbasin has one set of watershed characteristics that are essentially assumed
to be uniform over each subbasin.
• The infiltration/watershed loss and rainfall excess are calculated for each sub-
basin independent of the other sub-basins.
• All such sub-responses are routed through the channel to the watershed outlet,
thus yielding an overall watershed response.
• Examples of semi-distributed models are:
– Agricultural Non-Point Source Pollution Models (AGNPS-98) (Ronald et al., 1999),
– Areal Nonpoint Source Watershed Env.Simulation (ANSWERS) (Dillaha et al., 2007)
– Environmental Policy Integrated Climate (EPIC) model (Williams et al ., 2009),
– Hydrologic Simulation Program-Fortran (HSPF) (Donigian et al ., 2007),
– Soil and Water Assessment Tool (SWAT) (Arnold et al. 1998)
 Distributed Models
• Distributed models are an extension of the lumped and semi-distributed
models.
• Distributed models attempt to simulate both the spatial heterogeneity and
the physical processes occurring within a watershed.
• The distributed models divide the basin into elementary unit areas such
as grid cells and solve basic physical equations to simulate the watershed
processes.
• In such models, the flows are routed from one grid cell to another as
water drains through the basin.
• This allows the heterogeneity of the watershed to be simulated at each of
the grid cells.
• Grid resolution is generally chosen in such a way that it is small enough
to represent the spatial variation of major runoff processes such as
rainfall, infiltration, evapotranpiration, etc., but large enough to be
practically computational (Paudel, 2010).
 Contd., Distributed Hydrologic Models
• Distributed models take into account the spatial variability of
hydrologic variables for a given watershed as well as the hydrologic
response at ungauged locations within the basin.

• The examples of distributed models are

– 3D-Catchment Conceptual Multilayer Model Kinematic Runoff and
Erosion Model (KINEROS2) (Carl et al ., 1999),
– Bochum Water Balance Model (BWBM) (David, 2000),
– Central Valley Groundwater and Surface Water Model (C2VGSM) (Negil
et al ., 2000),
– Dynamic Watershed Simulation Model (DWSM) (Deva et al ., 2000)
– Hydrological River Basin Environment Assessment Model
(HydroBEAM) (Toshiharu et al ., 2007),
– Integrated Water Flow Model (IWFM) (Emin et al ., 2008),
– Storm Water Management Model (SWMM) ( Lewis et al ., 2007)
– Varaible Infiltration Capacity (VIC) model
 Lumped Vs Semi-Distributed HMs
• The lumped hydrologic models • The main advantage of semi-
were widely used in several distributed models is that their
applications because of their structure is more representative of
simple structure, minimum data physical hydrological process than the
requirements, fast set up and structure of lumped models, and that
calibration, and easy use. these models are less demanding on
• The representation of hydrologic input data than fully distributed
processes in lumped hydrologic models.
models is usually very • Semi-distributed model is a trade-off
simplified; between fully distributed and lumped
• Key limitation of simple lumped models.
models is that their parameters • The SWAT model is a popular model
are not directly related to the which has several practical
physical characteristics of the applications for long-term simulations
catchment. in watersheds dominated by
agricultural land use
 Soil and Water Assessment Tool (SWAT)
• SWAT is a physically based semi distributed parameter model which have been
developed to predict runoff, erosion, sediment and nutrient transport from
agricultural watersheds under different management practices (Arnold et al., 1998).

• Applicability: SWAT is most versatile model and has been widely used in various
regions and climatic conditions

• Interface: SWAT can be linked to a GIS system through an interface that enables
spatial data processing and visualization easy.

• Time scale: The model can simulate long periods, up to several years operating with
a daily time step.

• Data organization : The watershed is divided into a number of sub-basins. These

sub basins are discretized into HRUs, which are lumped land areas that are
comprised of unique land cover and soil combinations and parameters are calculated
for each HRU.
 Fig source :http://swat.tamu.edu/documentation/2012-io
t
SWt  SWo  (R
t 1
i
day
 Qsurf
i
 ETai  Wseep
i
 Qgw
i
)
 Surface runoff
Curve Number (SCS-CN) method :
i
( Rday  I a )2  100 
Qsurf  i S  25.4   10 
( Rday  I a  S )  CN 
Assumption that the ratio of actual runoff to potential runoff is the
same as the ratio of actual retention to potential retention.

where Qsurf is the accumulated runoff (mm), Riday is the rainfall depth of the day (mm), Ia is the
initial abstractions which includes surface storage, interception and infiltration prior to runoff
(mm), and S is the retention parameter (mm). The retention parameter varies spatially due to
changes in soils, land use, management and slope and temporally due to changes in soil water
content.
i
(Rday  0.2S )2
Qsurf  i
(Rday  0.8S)

CN is the curve number for the day.

The initial abstraction Ia is estimated as 0.2S
Runoff will occur when Rday > Ia .
 Surface runoff (contd..)
Characteristics Remarks ver
AMC I, II and III are considered 2000
 100  based on the last 5 days rainfall ver
S  25.4   10 
 CN  (SCS, 1971)

Replacing the five-day AMC with 2004

SMI 
 SMI  R i 1
day i 1
 Qday
a SMI (Williams and LaSeur
1976; Williams et al 1984)
ver

1.0  b  SMI  R  LE
i 1
day i 1

S is a function of soil moisture 2005

 SW  content and water holding 2009
S t  S max  1   capacity (Arnold et al. 2000) ver
 SW  exp( w1  SW .w 2) 

where S is the retention parameter for a given day (mm),

Smax is the maximum value the retention parameter can achieve on any given day (mm),
SW is the soil water content of the entire profile excluding the amount of water held in the
profile at wilting point (mm), and
w1 and w2 are shape coefficients. LE = lake evaporation (mm)
SMI = soil-moisture index, b = soil-moisture depletion coefficient,
 Sediment Yield
• SWAT uses the Modified Universal Soil Loss Equation (MUSLE)
• Sediment generation from each HRU is calculated by the following equation:

SY  11.8  (Qsurf  q peak  areahru )0.56  Kusle  Cusle  Pusle  LSusle  CFRG

where, SY is the sediment yield on a given day in metric tons, Qsurf is the surface runoff from the watershed in mm/ha,
qpeak is the peak runoff rate in cubic meter per second, areahru is the area of HRU,
KUSLE is the USLE soil erodability factor, CUSLE is the USLE land cover and management factor,
PUSLE is the USLE support practice factor, LSUSLE is the USLE topographic factor, a
CFRG is the coarse fragment factor. qpeak is the peak runoff rate in cubic meters per second,
C is the runoff coefficient, which is the ratio of inflow rate to peak discharge rate,
I is the rainfall intensity in mm/hr, and A is the sub-basin area in hectare. Rday is the rainfall for the day in mm

CIA Q surf
q peak  C 
360 R day

• Advantages:
– Elimination of dependency on soil physical factors and sediment delivery ratios.
– As the runoff factor reflects energy used in sediment detachment as well as transport.
95
 Disadvantages of
Conceptual and Physically based Models
• Data Intensive
• Complex and large number of parameters are involved
• Implementation and calibration of such a model can typically
present various difficulties
• Requiring sophisticated mathematical tools, significant
amounts of calibration data and some degree of expertise and
experience with the model.
• Data on the physical characteristics of the catchment at the
grid scale is required
Hydrologic Design
-Reservoir / Dams
- Capacity Estimation
 Fixing the Capacity of Reservoir
• Annual demand of water at a particular site may be less than the
total inflow, but the time distribution of demand may not match
the time distribution of inflows resulting in surplus in some
periods and deficit in some other periods.
• Reservoir
– It is a storage structure that stores water in periods of excess flow (over
demand) in order to supply water during non-monsoon season.
– It enables a smooth regulation of the available water to best meet the
specified demands.

• Reservoir Sizing:
– It involves determination of the required storage capacity of the reservoir
when inflows and demands in a sequence of periods are given.
 Reservoir Storage Capacity Components
• Flood storage capacity reserved to
reduce potential downstream flood
damage during flood events.
• Active storage used for downstream
flow regulation and for water
supply, recreational development or
hydropower production (for
conservation purposes).
• Dead storage required for sediment
collection

Fig: Zones of storages in a reservoir

• Reservoir sizing deals with determination of Active Storage requirement.

• Inflows are assumed to be deterministic
 Mass Diagram Method
• Developed by W. Rippl (1883).

• It involves finding the maximum positive cumulative difference

between a sequence of pre-specified (desired) reservoir releases Rt
and known inflows Qt.

• This can be visualized as starting with a full reservoir, and going

through a sequence of simulations in which the inflows and releases
are added and subtracted from that initial storage volume value.

• Doing this over two cycles of the record of inflows will identify the
maximum deficit volume associated with those inflows and
releases. This is the required reservoir storage.
 I, D
Ripple diagram /
b
Mass curve method a
D
a+b

t1 t2 t
∑It, ∑Dt ∑Dt

• Condition to be satisfied?? A
Sum of all the desired releases a+b b
a ∑ It
should not exceed the sum of all B
the inflows over the same
sequence of time periods. t
t1 t2

• The difference between the lines (a+b) tangent to the demand line
(∑Dt) drawn at the highest and lowest points (A and B, respectively)
of mass curve (∑ It) gives the rate of withdrawal from reservoir
during that critical period.
• The maximum cumulative value between tangents is the required
storage capacity (active storage).
 Fixing the Capacity of a reservoir
Required Storage = Dt- St (over a year)
 Example: Rippl’s Mass Diagram Method
Given: inflows for a nine period sequence are 1, 3, 3, 5, 8, 6, 7, 2 1.
Estimate the required reservoir storage capacities for two different
release rates of 4 and 3 units
 Alternative Method for Identifying Reservoir Capacity

• The active storage capacity, Ka, will be equal to Analytical equivalent of graphical method
the maximum accumulated storage deficit one  j 
K a  max .    Rt  Qt   ,
can find over some interval of time within two  t i 
successive record periods, T.
where 1  i  j  2T

• The Mass diagram method does not account for losses.

• The equivalent to this kind of procedure is known as the sequent peak method
 Solve and find the capacity of the reservoir
required??

Mean flow Cumulative Cumulative

Month cu.m/d Volume/month Demand flow demand

Jan 60 6696000 4464000 6696000 4464000

Feb 45 4536000 4032000 11232000 8496000

mar 35 3906000 4464000 15138000 12960000

Apr 25 2700000 4320000 17838000 17280000

May 15 1674000 4464000 19512000 21744000

Jun 22 2376000 4320000 21888000 26064000

Jul 50 5580000 4464000 27468000 30528000

Aug 80 8928000 4464000 36396000 34992000

Sep 105 11340000 4320000 47736000 39312000

Oct 90 10044000 4464000 57780000 43776000

Nov 80 8640000 4320000 66420000 48096000

Dec 70 7812000 4464000 74232000 52560000

 80000000.00

Cum. Inflow
60000000.00 Cum. Demand
Cum. Inflow/demand

40000000.00

20000000.00

0.00

0.00 4.00 8.00 12.00

Time
 Mean
flow
cu. Volume/ Cumulative Cumulative
Month m/d month Demand Surplus Deficit surplus deficit

Jan 60 6696000 4464000 2232000 2232000

Feb 45 4536000 4032000 504000 2736000

mar 35 3906000 4464000 -558000 -558000

Apr 25 2700000 4320000 -1620000 -2178000

May 15 1674000 4464000 -2790000 -4968000

Jun 22 2376000 4320000 -1944000 -6912000

Jul 50 5580000 4464000 1116000 1116000

Aug 80 8928000 4464000 4464000 5580000

Sep 105 11340000 4320000 7020000 12600000

Oct 90 10044000 4464000 5580000 18180000

Nov 80 8640000 4320000 4320000 22500000

Dec 70 7812000 4464000 3348000 25848000

 Sequent Peak Algorithm
• If N year flow data are available
• It is assumed the cyclic pattern repeats
700

600

500

400

300

200

100

0
0 100 200 300 400 500
T i me i n mont hs
T i me s e r i e s pl ot of T ot a l i nf l ow
 Sequent Peak Algorithm
• Estimate the Net-flow volume
= Inflow-outflow (demand)
• Plot the cumulative net-flow Vs time
• Find the sequent peak (high) and trough (low)
• The maximum difference gives the capacity of the reservoir.
 Sequent Peak Algorithm
• Let the following notation
– Kt be the maximum total storage requirement needed for periods 1
through period t.
– Rt be the required release in period t, and
– Qt be the inflow in that period.

• Set K0 =0, the SPA procedure involves calculating Kt using the

following equation (for upto twice the total length of record).
K t  Rt  Qt  K t 1 , if Positive
t  1,2,..., 2T 
 0, otherwise

K a  Max  K t , t  1, 2,..., 2T 

• Thus, the required storage capacity Ka is equal to the maximum

of all Kt for the specified releases Rt and inflows Qt.
 Sequent Peak Analyses

• Inflows for a nine period sequence

are 1, 3, 3, 5, 8, 6, 7, 2 and 1.

• Constant release required, Rt=3.5

• This method does not require all the

releases to be same.

• Stopping Criteria
– Kt value repeats for the
corresponding period OR
– Twice the number of periods
 Drought
• Meteorological drought
– If rainfall is decreased more than 25% of normal rainfall
• Severe drought – if rainfall is 75% of normal rainfall
• Severe drought – if rainfall is 40-75% of normal rainfall

• Hydrological drought
– Depletion of surface and ground water
– Drying of lakes reservoirs aquifers

• Agricultural drought
– When soil moisture and rainfall are inadequate during growing period of a
crop

• Economic drought
 Self Study
• Solve all the problems given in this section

• Others techniques of stream flow measurement

– Electromagnetic method
– Ultrasonic method
– Indirect methods
• Flow measuring structures (weirs and notches?)
• Slope area methods (based on resistance equation for
uniform flow and Manning’s formula)
 Current research in this area
1. Application of ANN and GP in rating
curve parameter estimation
2. Finding new sophisticated instruments
for stream-flow measurement
3. Etc…..
 Research area in this topic
• Establishment of rainfall-runoff relationship
– Black box models –ANN
– Conceptual models
– Physical models
• Inclusion of GIS and RS
• Drought prediction and mitigation