Ministry of Agriculture

Participatory Small Scale Irrigation Development Program

Training Manual on Hydrology

Bahir Dar University

Institute of Technology
School of Civil & Water Resources Engineering

October, 2013

Bahir Dar, Ethiopia

Bahir Dar University, ioT School of Civil & Water Resources Engineering

Table of Contents
1. INTRODUCTION TO HYDROLOGY ........................................................... 1
1.1 Introduction ..................................................................................................... 1
1.2 Hydrologic Cycle............................................................................................. 1
1.3 System concept .............................................................................................. 3
1.4 Hydrologic model classification ................................................................ 5
2 BASIC HYDROLOGIC PRINCIPLES.......................................................... 7
2.1 Introduction ..................................................................................................... 7
2.2 Variability of Precipitation .......................................................................... 7
2.3 Areal Averaging of Precipitation .............................................................. 8
2.3.1 Design storm .......................................................................................... 10
2.3.1.1 Design Precipitation Depth ........................................................ 10
2.3.1.2 Areal Precipitation Depth ........................................................... 10
2.3.1.3 Design hyetograph from storm event analysis ................. 11
2.4 Evaporation .................................................................................................... 14
2.5 Catchment properties ................................................................................ 15
2.6 Runoff ............................................................................................................... 15
3 SURFACE WATER RESOURCE ASSESSMENT .................................... 20
3.1 Water Balance Study .................................................................................. 20
3.2 Water requirements and irrigable area............................................... 21
3.3 Water Availability Studies ........................................................................ 22
3.3.1 Flow Duration Analysis ....................................................................... 23
3.3.2 Low Flow Frequency Analysis .......................................................... 25
3.4 Ungauged Catchments .............................................................................. 27
3.5 Demand-Water Supply Analysis ............................................................ 29
4 DESIGN FLOOD STUDIES ........................................................................... 30
4.1 Introduction ................................................................................................... 30
4.2 Flood Frequency Analysis ......................................................................... 30
4.2.1 Flood probabilities ................................................................................ 30
4.2.2 Return Period and Risk....................................................................... 32
4.2.3 Probability Distributions..................................................................... 32
4.2.4 Fitting theoretical distributions to sample data ....................... 33
4.2.5 The Graphical Approach..................................................................... 33
4.2.6 Computation of sample statistics: ................................................. 35
4.2.7 Extreme Value Type I or Gumbel’s Distribution: ..................... 35
4.2.8 Log-Pearson Type III Distribution ................................................. 38
4.2.9 Instantaneous versus Daily Peaks ................................................. 41
4.3 Un-gauged Catchments............................................................................. 44
4.3.1 Regional flood frequency estimation ............................................ 44
4.3.2 Flood marks and interviews ............................................................. 44

i
Bahir Dar University, ioT School of Civil & Water Resources Engineering

4.3.3 Using Regional Formulae ................................................................... 45

4.3.4 Rainfall Runoff Methods ..................................................................... 45
4.3.4.1 The Rational Method .................................................................... 46
4.3.4.2 The SCS Method ............................................................................ 48
5 FLOW ROUTING ............................................................................................... 56
5.1 Methods of flood routing: ......................................................................... 56
5.2 Applications of flood routing: .................................................................. 57
5.3 Hydrologic Routing Techniques .............................................................. 57
5.3.1 Modified puls Reservoir Routing ..................................................... 57
5.3.2 Muskingum Method of Channel Routing ...................................... 62

ii
Bahir Dar University, ioT School of Civil & Water Resources Engineering

1. Introduction to Hydrology
1.1 Introduction

Hydrology, which treats all phases of the earth's water, is a subject of great importance
for people and their environment. Practical applications of hydrology are found in such
tasks as the design and operation of hydraulic structures, water supply, wastewater
treatment and disposal, irrigation, drainage, hydropower generation, flood control,
navigation, erosion and sediment control, salinity control, pollution abatement,
recreational use of water and fish and wildlife protection

The hydrosciences deal with the waters of the earth: their distribution and circulation,
their physical and chemical properties, and their interaction with the environment,
including interaction with living things and, in particular, human beings. Hydrology may
be considered to encompass all the hydrosciences, or defined more strictly as the study of
the hydrologic cycle, that is, the endless circulation of water between the earth and its
atmosphere. Hydrologic knowledge is applied to the use and control of water resources
on the land areas of the earth; ocean waters are the domain of ocean engineering and the
marine Sciences.

1.2 Hydrologic Cycle

The hydrologic cycle is the central focus of hydrology. The cycle has no beginning or
end, and its many processes occur continuously. As shown schematicatty in-fig. 1.l, water
evaporates from the oceans and the land surface to become part of the atmosphere; water
vapor is transported and lifted in the atmosphere until it condenses and precipitates on the
land or the oceans; precipitated water may be intercepted by vegetation, become overland
fow over the ground surface, Infiltrate into the ground, flow through the soil as
subsurface flow, and discharge into streams as surface runoff. Much of the intercepted
water and surface runoff returns to the atmosphere through evaporation. The infiltrated
water may percolate deeper to recharge groundwater, later emerging in springs or seeping
into streams to form surface runoff, and finally flowing out to the sea or evaporating into
the atmosphere as the hydrologic cycle continues.

Seifu A. 1
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Figure 1-1 Gobal Hydrologic Cycle

Seifu A. 2
Bahir Dar University, ioT School of Civil & Water Resources Engineering

The hydrology of a region is determined by its weather patterns and by physical factors
such as topography, geology and vegetation. Also as civilization progresses, human
activities gradually encroach on the natural water environment, altering the dynamic
equilibrium of the hydrologic cycle and initiating new processes and events. For example,
it has been theorized that because of the burning of fossil fuels, the amount of carbon
dioxide in the atmosphere is increasing. This could result in a warming of the earth and
have far-reaching effects on global hydrology.

1.3 System concept

Hydrologic phenomena are extremely complex, and may never be fully understood.
However, in the absence of perfect knowledge, they may be represented in a simplified
way by means of the systems concept. A system is a set of connected parts that form a
whole. The hydrologic cycle may be treated as a system whose components are
precipitation, evaporation, runoff, and other phases of the hydrologic cycle. These
components can be grouped into subsystems of the overall cycle; to analyze the total
system, the simpler subsystems can be treated separately and the results combined
according to the interactions between the subsystems.

Precipitation Evaporation
Atmospheric

Σ
water

Interception

Transpiration
Surface
water

Overland flow Surface Runoff to

flow streams and
ocean

Infiltration Subsurface
flow
Subsurface

Σ
water

Groundwater Groundwater
recharge flow

Figure 1-2 The global hydrologic cycle is represented as a system.

Seifu A. 3
Bahir Dar University, ioT School of Civil & Water Resources Engineering

For most practical problems, only a few processes of the hydrologic cycle are considered
at a time and then only considering a small portion of the earth's surface. A more
restricted system definition than the global hydrologic system is appropriate for such
treatment; and is developed from a concept of the control Volume.

The control volume provides the framework for applying the laws of conservation of
mass and energy and Newton's second law to obtain practical equations of motion. In
developing these equations, it is not necessary to know the precise flow pattern inside the
control volume. What must be known are the properties of the fluid flow at the control
surface, the boundary of the control volume. The fluid inside the control volume is
treated as a mass; which may be represented as being concentrated at one point in space
when considering the action of external forces such as gravity.

A hydrologic system is defined as a structure or volume in space, surrounded by a

boundary, that accepts water and other inputs, operates on them internally, and produces
them as outputs (Fig, 1-3). The structure (for surface or subsurface flow) or volume in
space (for atmospheric moisture flow) is the totality of the flow paths through which the
water may pass as throughput from the point it enters the system to the point it leaves.
The boundary is a continuous surface defined in three dimensions enclosing the volume
or structure. A working medium enters the system as input, interacts with the structure
and other media, and leaves as output. Physical, chemical, and biological processes
operate on the working media within the system; the most common working media
involved in hydrologic analysis are water, air, and heat energy.

Figure 1-3 Schematic representation of system operation

By using the system concept, effort is directed to the construction of a model relating
inputs and outputs rather than to the extremely difficult task of exact representation of the
system details, which may not be significant from a practical point of view or may not be
known. Nevertheless, knowledge of the physical system helps in developing a good
model and verifying its accuracy.

The objective of hydrologic system analysis is to study the system operation and predict
its output. A hydrologic system model is an approximation of the actual system; its inputs
and outputs are measurable hydrologic variables and its structure is a set of equations
linking the inputs and outputs.

Seifu A. 4
Bahir Dar University, ioT School of Civil & Water Resources Engineering

1.4 Hydrologic model classification

Hydrologic models may be divided into two categories: physical models and abstract
models. Physical models include scale models which represent the system on a reduced
scale such as lysimeters, rainfall simulators, and; analog model which use another
physical system having properties similar to those the prototype, such as the flow of
electrical current to represent the flow of water.

Abstract models represent the system in mathematical form. In engineering hydrology,

four types of mathematical models are in current use:
1. Deterministic: formulated by using laws of physical or chemical
processes
2. Probabilistic: whether statistical or stochastic, governed by laws of
chance or probability. Statistical models deals with observed samples,
whereas stochastic models focus on the random properties of certain
hydrologic time series.
3. Conceptual: simplified representation of the physical processes, obtained
by lumping spatial and /or temporal variations, described in terms of
either ordinarily differential equation or algebraic equation.
4. Parametric: represents hydrologic processes by means of algebraic
equation that contain key parameters to be determined by empirical
means.

In principle, deterministic models mimic physical processes and should, therefore, be

closest to reality. In practice, however, the inherent complexity of physical phenomena
generally limits the deterministic approach to well defined cases for which a clear cause-
effect relationship can be demonstrated. Probabilistic methods are used to fit measured
data (i.e., statistical hydrology) and to model random components (stochastic hydrology)
in cases where their presence is readily apparent. Where simplicity is desired, conceptual
and parametric methods and models continue to play an important role in hydrologic
engineering practice.

The model may be located in a "tree" according to these choices, as shown in Fig. 1.4. 1.

Seifu A. 5
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Figure 1-4 Classification of hydrologic modes according to the way they treat the randomness and
space and time variability of hydrologic phenomenon

Seifu A. 6
Bahir Dar University, ioT School of Civil & Water Resources Engineering

2 Basic hydrologic principles

2.1 Introduction
Hydrology takes a quantitative view of the hydrologic cycle. Generally, equations are
used to describe the interaction between the various phases of the hydrologic cycle. The
following basic equation relates precipitation and surface runoff:
Q = P-L
In which Q = surface runoff, P = precipitation; and L = losses, or hydrologic abstractions.

Generally, the catchment has an abstractive capacity that acts to reduce total rainfall into
effective rainfall. The difference between total rainfall and effective rainfall is the losses
or hydrologic abstractions. The abstractive capacity is a characteristic of the catchment,
varying with its level of stored moisture. Hydrologic abstractions include interception,
infiltration, surface storage, evaporation, evapotranspiration. The difference between total
rainfall and hydrologic abstractions is called runoff. Therefore, the concept of effective
rainfall and runoff are equivalent.

2.2 Variability of Precipitation

Precipitation varies geographically, temporally and seasonally. Both regional and

seasonal variations of precipitation are very important for water resources planning and
hydrologic studies. Temporal variations in rainfall intensity are extremely important in
the rainfall-runoff process in urban areas. This is because urban areas have a very short
response time due to the large impervious areas and artificial drainage network.

In Ethiopia, the seasons and rainfall regimes are classified based on mean annual and
mean monthly rainfall distribution. There are main three rainfall regimes in Ethiopia
(Bekele, 1997). These three rainfall regimes are delineated as:
• Mono-Modal (Single maxima)
• Bi-modal type-1 (Quasi-double maxima)
• Bi-modal type-2 (Double maxima)
As shown in Fig 2.1 below, Mono-modal, Bi-modal type-1 and Bi-modal type-2 are
designated by letter B, A and C, respectively.

Seifu A. 7
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Figure 2-1 Rainfall Regimes of Ethiopia (Bekele. 1997)

The area designated as region B in Fig 2.1 is dominated by single maxima rainfall
pattern. However, the wet period decreases northwards from about ten month in the south
west to only about four month in the north. Thus region-B is sub-divided into three parts
designated as b1, b2 and b3, where the wet period runs from February/March to
October/November, April/May to October/November and from June/July to
August/September, respectively.

The area designated as region A is characterized by quasi-double maxima rainfall pattern,

with a small peak in April and maximum peak in August. The central and most of the
eastern half of the country is included in this rainfall regime. The two rainy periods are
locally known as ‘Kiremt’ (June to September) and ‘Belg’ (February to May), which are
the long and short rainy periods, respectively. Short ‘dry’ period, which covers the rest of
the year (i.e. October to January), is known as Bega.

The area identified as region C in the Fig 2.1 is dominated by double maxima rainfall
pattern with peak during April and October. The southern and the southeastern parts of
the Ethiopia are included in this rainfall regime. Two rainy periods are from March to
May and from September to November. Two dry periods are from June to August and
from December to February.

2.3 Areal Averaging of Precipitation

Seifu A. 8
Bahir Dar University, ioT School of Civil & Water Resources Engineering

When a drainage basin or an area has more than one raingage in or near the area
considered, these rain gages inevitably record different amounts of precipitation, whether
it is for a single rainstorm or over a specified period of time. Often one needs to find the
average precipitation over the area. Three commonly used methods are described:

1. Arthiemetic mean method

The areal average precipitation, is computed using

P1 + P2 + .....+ Pn 1 n
Pav = = ∑Pi
n n i=1
n = total number of precipitation stations the drainage basin;
Pj = precipitation for the j-th station.
This method applies when the stations are fairly evenly distributed and the variation of
areal distribution of precipitation is uniform.

2. Thiessen Polygon method

The locations of the rain gages are plotted on a scale map of the catchment and
surrounding area. The locations (stations) are joined with straight lines in order to form a
pattern of triangles. Perpendicular bisectors to the sides of these triangles are drawn to
enclose each stations within a polygon called thiessen polygon, circumscribing an area of
influence. The average precipitation over the catchment is calculated by weighing each
station’s rainfall depth in proportion to its area of influence, i.e., if there are J gages, and
the area within the catchment assigned to each is Aj, Pj is the rainfall recorded at the jth
gage, the areal average precipitation for the catchment is:
1 J
Pav = ∑ Aj Pj
A j=1
There is a tutorial session for this to show how GIS can be used to estimate average
precipitation values. A separate document will be attached during the tutorial.

3. Isohyetal method

The locations of the rain gages are plotted on a scale map of the catchment and
surrounding area. Each station’s rainfall depth is used to draw isohyets through the
catchment in a manner similar to that used in the preparation of topographic contour
maps. Once the isohyetal map is constructed, the area Aj b/n each pair of isohyets, within
the catchment, is measured and multiplied by the average Pi of the rainfall depths of the
two boundary isohyets to compute the areal average precipitation by
1 J
Pav = ∑ Aj Pj
A j=1

Seifu A. 9
Bahir Dar University, ioT School of Civil & Water Resources Engineering

2.3.1 Design storm

A design storm is a precipitation pattern defined for use in the design of a hydrologic
system. Usually the design storm serves as the system input, and the resulting rates of
flow through the system are calculated using rainfall-runoff and flow routing procedures.
A design storm can be defined by a value for precipitation depth at a point, by a design
hyetograph specifying the time distribution of precipitation during a storm, or by an
isohyetal map specifying the spatial pattern of the precipitation.

Design storms can be based upon historical precipitation data at a site or can be
constructed using the general characteristics of precipitation in the surrounding region.
Their application ranges from the use of point precipitation values in the rational method
for determining peak flow rates in storm sewers and highway culverts to the use of storm
hyetographs as inputs for rainfall-runoff analysis of urban detention basins or for spillway
design in large reservoir projects.

2.3.1.1 Design Precipitation Depth

Point precipitation is precipitation occurring at a single point in space as opposed to areal

precipitation which is precipitation over a region. The point depth appies for all areas less
than a point area, smallest area below which variations of storm depth with catchment
area can be assumed to be nefigible. In US, the point area is usually taken as 25km2. For
point precipitation frequency analysis, the annual maximum precipitation for a given
duration is selected for each year of historical record. For each duration, frequency
analysis is performed on the data.

2.3.1.2 Areal Precipitation Depth

Frequency analysis of precipitation over an area has not been as well developed as has
analysis of point precipitation. In the absence of information on the true probability
distribution of areal precipitation, point precipitation estimates are usually extended to
develop an average precipitation depth over an area. The areal estimate may be either
storm-centered or location-fixed. For the location-fixed case, one accounts for the fact
that precipitation stations are sometimes near the storm center, sometimes on the outer
edges, and sometimes in-between the two. An averaging process results in location-fixed
depth-area curves relating areal precipitation to point measurements. Fig. 2.2 provides
curves for calculating areal depths as a percentage of point precipitation values (World
Meteorological Organization, 1983). Area to point ratio for Ethiopia is shown in Table 2-
1.

Seifu A. 10
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Figure 2-2 Depth area curves for reducing point rainfall to obtain areal average value.(chow, 1988)

2.3.1.3 Design hyetograph from storm event analysis

By analysis observed storm events, the time sequence of precipitation in typical storms
can be determined. The U.S. Department of Agriculture, Soil Conservation Service
(1986) developed Synthetic storm hyetograph for use in US for storms of 6 and 24 hours
duration. Table 2-2 presents the cumulative hyetographs. Similarly, figure 2-3 shows a
similar case for Ethiopia developed by Continental/CES consultants. For the
determination of time increment, D, one hour will be used for the first most intensive 6
hour. However, for rapid concentration time (Tc less than 3 hours), incremental time D
will be one-half hour. For Tc values of 10 to 15 hours, a time increment D of 2 hour is
recommended.

The time period, D, may be lengthened in the later part of the storm to reduce
computations. This will result in a poorer definition of the recession limb of the
hydrograph, but it has little effect on the design.

Seifu A. 11
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Figure 2-2 Area to point ratio (%) (Continental/CES, SSIP, Handbook, 1997)

Area Duration 0.5 1 2 3 4 5 6 9 12 15 18 21 24

2
(km ) (hrs)
25 88 78 82 85 87 88 88 91 92 93 93 94 94
50 61 71 78 82 84 85 87 89 90 91 92 92 93
75 57 67 75 79 82 84 83 87 89 90 91 91 92
100 54 65 73 78 80 82 83 86 88 89 90 91 91
125 52 63 72 76 79 81 82 85 87 88 89 90 91
150 50 61 70 75 78 80 81 84 86 88 89 89 90
175 48 59 69 74 77 79 81 84 86 87 88 89 90
200 46 58 68 73 76 78 80 83 85 87 88 88 89
225 45 57 67 72 75 77 78 82 85 86 87 88 89
250 44 55 66 71 74 77 78 82 84 86 87 88 88
275 42 54 65 70 74 76 78 81 84 85 86 87 88
300 41 53 64 70 73 75 77 81 83 85 86 87 88
325 40 53 63 68 72 75 77 80 83 84 86 87 87
350 38 52 63 68 72 74 76 80 82 84 85 86 87
375 39 51 62 68 71 74 76 80 82 84 85 86 87
400 38 50 61 67 71 73 75 79 82 83 85 86 87
425 37 50 61 67 70 73 75 79 81 83 84 85 86
450 36 49 60 66 70 72 74 79 81 83 84 85 86
475 36 48 60 66 69 72 74 78 81 83 84 85 86
500 35 48 59 66 69 72 74 78 80 82 84 85 86
525 34 47 59 65 68 71 73 78 80 82 83 85 85
550 34 47 58 64 68 71 73 77 80 82 83 84 85
575 33 46 58 64 68 71 73 77 80 82 83 84 85
600 33 45 57 63 67 70 72 77 79 81 83 84 85
625 32 45 57 63 67 70 72 76 79 81 83 84 85
650 32 45 56 63 67 69 72 76 79 81 82 84 84
675 31 45 56 62 66 69 71 76 79 81 82 83 84
700 31 45 56 62 66 69 71 76 78 80 82 83 84
725 31 45 55 62 66 69 71 75 78 80 82 83 84
750 30 45 55 61 65 68 71 75 78 80 82 83 84

Seifu A. 12
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Table 2-2-2 SCS rainfall distribution (chow, 1988)

Time Distribution of Rainfall

100
Rainfall ratio as % of Daily

90
80

70
Rainfal

60
50
40
30
20
10
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Time (hr)

Figure 2-2-3Graph giving time distribution of rainfall (Continental/CES, SSIP, Handbook, 1997)

Seifu A. 13
Bahir Dar University, ioT School of Civil & Water Resources Engineering

2.4 Evaporation

The two main factors influencing evaporation from an open water surface are the supply of energy to
provide the latent heat of vaporization and the ability to transport the vapor away from the evaporative
surface. Solar radiation is the main source of heat energy. The ability to transport vapor away from the
evaporative surface depends on the wind velocity over the surface and the specific humidity gradient
in the air above it. Evaporation from the land surface comprises evaporation directly from the soil and
vegetation surface, and transpiration through plant leaves, in which water is extracted by the plant's
roots, transported upwards through its stem, and diffused into the atmosphere through tiny openings in
the leaves called stomata. The processes of evaporation from the land surface and transpiration from
vegetation are collectively termed evapo-transpiration. Evapo-transpiration is influenced by the two
factors described previously for open water evaporation, and also by a third factor, the supply of
moisture at the evaporative surface. The potential evapo-transpiration is the evapotranspiration that
would occur from a well vegetated surface when moisture supply is not limiting, and this is calculated
in a way similar to that for open water evaporation. Actual evapo-transpintion drops below its
potential level as the soil dries out.

Evaporation may be computed by the aerodynamic method when energy supply is not limiting and by
the energy balance method when vapor transport is not limiting. But, normally, both of these factors
are limiting, so a combination of the two methods is needed.
a) Latent heat of vaporization(Lu), J/kg
Lu=2.5*10^6-2370Tave
b) Radiation term , Er
Er=Rn/ (Lu*ρw) where ρw is density of water
c) Aerodynamic term, (Ea)
Ea=B(es-ea) (mm/day)
Where:
B =0.0027*(1+U/100) (mm.pa/Day) where U is the 24 hour wind run in km/day measured at a
height of 2m above the ground.

ETo (mm/day) = ∆/(∆+ϒ) *Er + ϒ/(∆+ϒ)*Ea

The FAO Penman-Monteith method is the sole standard alternative method for the computation of
ETo from meteorological data. This equation determines the evapotranspiration from the hypothetical
grass reference surface and provides a standard to which evapotranspiration in different periods of the
year or in other regions can be compared and to which the evapotranspiration from other crops can be
related. This method will be more demonstrated with excel in tutorial session. A separate
document will be distributed.

Estimation of ET in areas, where adequate meteorological data are not available, is one of the
challenges faced by water resource managers. Hence, a simplified approach, which is less data
intensive, is crucial. A new simple temperature method (Enku and Melesse, 2013) is developed by

Seifu A. 14
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Temesgen Enku (member of the school of Civil & Water Resources Engineering, Bahir Dar
University), which uses only maximum temperature data to estimate ET.

The new simple empirical temperature method, which we named ‘Enku’s simple temperature method’,
was developed as follows:

where ETo is the reference evapotranspiration (mm/day); n = 2.5, which can be calibrated for local
conditions; k =48*Tmm-330 for combined wet and dry conditions.

2.5 Catchment properties

The watershed is the basic unit of all hydrologic analysis and designs. Any watershed can be
subdivided in to a set of smaller watersheds. Usually a watershed is defined for a given drainage point.
This point is usually the location at which the analysis is being made and is referred to as the
watershed “outlet”. The watershed, therefore, consists of all the land area that drains water to the
outlet during a rainstorm. The boundary of a watershed therefore consists of the line drown across the
contours joining the highest elevations surrounding the basin. Watersheds considered in engineering
hydrology vary in size from a few hectares in urban areas to several thousand square kilometers for
large river basins.

A common task in hydrology is to delineate a watershed from a topographic map. In recent years the
use of GIS (Geographic Information System) has become increasingly popular and has facilitated
much of the work of hydrologists. The use of DEMs (Digital Elevation Models) in particular has made
watershed delineation a relatively smooth procedure. To demonstrate how to use GIS, there is a
tutorial and a separate document to be distributed during the training.

2.6 Runoff

Surface runoff, or simply runoff, refers to all waters flowing on the surface of the earth, either by
overland flow or by channel flow in rills, gullies, streams, or rivers. Surface runoff is a continuous
process by which water is constantly flowing from higher to lower elevations by the action of
gravitational forces. Small streams combine to form larger streams which eventually grow into rivers.
In time, rivers carry their flow into the ocean, completing the hydrologic cycle.

Runoff consists of water from three sources: (1) surface flow, (2) interflow, and (3) groundwater flow.
Surface flow is the product of effective rainfall, i.e., total rainfall minus hydrologic abstractions.
Surface flow is also called direct runoff. Direct runoff is responsible for flood flows

Interflow is subsurface flow, i.e., flow that takes place in the unsaturated soil layers located beneath
the ground surface. Interflow consists of the lateral movement of water and moisture toward lower
elevations, and it includes some of the precipitation abstracted by infiltration. Ground water flow
takes place below the groundwater table in the form of saturated flow through water bearing

Seifu A. 15
Bahir Dar University, ioT School of Civil & Water Resources Engineering

formations located beneath the soil mantle. Both Interflow and groundwater flow are characteristically
a slow process.

Antecedent moisture

Surface runoff is directly related to effective precipitation, and effective precipitation is inversely
related to the hydrologic abstractions. During rainy periods, infiltration pays a major role in
abstracting total precipitation. Actual infiltration rate and amounts are highly dependent on the initial
level of soil moisture, which is defined as antecedent moisture.

The effect of infiltration rate and antecedent moisture on runoff is modeled widely by SCS runoff
curve number model. The SCS mode is based on a non-linear rainfall-runoff relation that includes a
third variable called the runoff curve number, or CN. The CN value is determined by a detailed
evaluation of soil type, vegetation and land use patterns, antecedent moisture, and hydrologic
condition of the catchmnet surface.

Figure 2-6 variables in SCS method

Q=
( P − 0 .2 S ) 2 P ≥ 0.2S
P + 0 .8 S

 1000 
S = 2.54 − 10 
 CN 

Curve numbers have been tabulated by the Soil Conservation on the basis of soil type and land use.
Four soil groups are defined:
Croup A: Deep sand, deep loess, aggregated silts
Group B: Shallow loess, sandy loam
Group C; Clay loams, shallow sandy loam, soils low in organic content, and soils usually high in clay
Group D: Soils that swell significantly when wet, heavy plastic clays. And certain saline soils

Seifu A. 16
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Table 2-3 Classification of antecedent moisture classes (AMC) for the SCS method of rainfall
abstractions
Growing Season Dormant Season
Antecedent Five-Day Five-Day
Condition Antecedent Rainfall Antecedent Rainfall

AMC I Less than 36 mm Less than 13 mm

AMC II 36 to 53 mm 13 to 28 mm

AMC III Over 53 mm Over 28 mm

Runoff concentration
An important characteristic of surface runoff is its concentration property. Runoff concentration
implies that the flow rate at the outlet will gradually increase until rainfall from the entire catchment
has had time to travel to the outlet and is contributing to the flow that point.

The time that it takes a parcel of water to travel from the farthest point in the divide to the catchment
outlet is referred to as the time of concentration, or concentration time.

In practice, concentrated and super-concentrated flows are typical of small catchments, i.e., those
likely to have short concentration times. On the other hand, sub-concentrated flows are typical of
midsize and large catchments, i.e., those with longer concentration times.

Seifu A. 17
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Fig 2-7 Runoff Concentration types of Catchment response: (a) concentrated; (b) Super concentrated; (c) Sub
concentrated

Seifu A. 18
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Table 2-4 Summary of time of concentration formulas.

Seifu A. 19
Bahir Dar University, ioT School of Civil & Water Resources Engineering

3 Surface Water Resource Assessment

Surface water resources are the water resources in streams, rivers and fresh water lakes. The
availability of water both in quantity and quality, and the temporal distribution of the resource are very
important in developing these resources for irrigation or other water uses. The availability of water in
rivers is studied using streamflow records in gauged catchments or can be predicted indirectly from
rainfall for ungauged catchments. Regionalized studies of river flows also enable to estimate water
availability in ungauged catchments. The assessment of the balance between the supply and the
demand will indicate the need for storage. This is done in what is commonly known as water balance
study. The following concepts are important:

Water Demand or Water Use is defined as water withdrawal for a specific purpose, such as
domestic use, irrigation, or industrial processing.

Agricultural water use: Water used for soil cultivation, crop production and livestock rearing.

In stream water use: Water that is used, but not withdrawn, from a surface-water or
groundwater source for such purposes as hydroelectric power generation, navigation, water-
quality improvement, fish propagation, and recreation. Sometimes called non-withdrawal use
or in-channel use.

Off stream water use: Water withdrawn or diverted from a ground- or surface-water source
for public-water supply, industry, irrigation, livestock, thermoelectric power generation and
other uses. Sometimes called off-channel use or withdrawal.

Irrigation: Artificial application of water on lands to assist in the growing of crops and
pastures or to maintain vegetative growth in recreational lands such as parks.

Irrigation using surface water sources can be carried out either using river diversion (gravity or by
pumping) or by means of dam reservoirs. In this presentation, focus will be given to small scale or
micro-scale irrigation by means of river diversion.

3.1 Water Balance Study

The water balance study is a very important item for planning any Water Resources Development
Project. It has to take into consideration the followings:

Availability of water to be based on historic data of flow taking into consideration
o The required dependability
o Proposed or earmarked u/s and d/s utilization

Demand for all purposes
o Irrigation
o Water supply for

Seifu A. 20
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Human consumption

Livestock

Industries

Environmental requirements for protecting ecology in
o Rivers
o Lakes

Losses due to
o Evaporation from reservoir
o Losses in irrigation system due to

Conveyance in canal system

Application in the field

3.2 Water requirements and irrigable area

Crops require a large amount of water for irrigation, and it is important to calculate water requirements
accurately, both to design the supply canal and the pump (if any), and to check that enough water is
available from the source. The amount of water required by a crop depends on the local environment,
the climate, the crop and its stage of growth, and the degree to which the crop may be stressed. This
requirement may be expressed as a uniform depth of water over the area in millimeters per day
(mm/d).

The monthly average meteorological parameters Maximum Temperature, Minimum Temperature,

Windspeed, Sunshine Hours and Relative humidity are used to compute the Reference Crop
Evapotranspiration (ETo). By taking the crop coefficients, Kc for the different types of crops, the crop
water requirement ETc can be computed:
ETc = Kc * ETo Eq. 3-1
Taking into account the total irrigable area, A and the different types of crops, the water requirement
on monthly basis or on a 10 day basis can be computed.

Hence, for a given crop & for each month or 10 day period, i
IRi = ETc – ER i Eq. 3-2
Where
IRi is the net irrigation water requirement for period i
ERi is the effective rainfall over the period i.
This IR is the net irrigation water requirement at the crop field. Additional water has to be supplied to
take account of field application losses which, with surface irrigation, are typically about 40 per cent,
giving an application efficiency of 0.60. The field irrigation requirement (IF) can be estimated as:
IF = IR/0.60 Eq. 3-3

Seifu A. 21
Bahir Dar University, ioT School of Civil & Water Resources Engineering

The required canal discharge depends on the water losses from the canal by seepage and evaporation.
This is taken into account as canal loss efficiency. For a design command area A (m2), the design
discharge required Q (l/s) for irrigation hours (H)every day, is given by the field-irrigation
requirement multiplied by the area, divided by the time (in seconds):
Q = IF x A /(Hx60x60) + canal losses Eq. 3-4

Figure 3-1 A schematic plan view of a typical small scale irrigation system

3.3 Water Availability Studies

Whenever it is aimed to use streamflows without regulating reservoir, the adequacy of streamflow to
meet requirements for irrigation, hydropower, domestic and industrial water supply has to be checked.
Furthermore, the question of quality of the environment often depends on the condition of river flows.
Thus, the connection between quantitative and qualitative aspects of water resources is sensitive
during low water periods.

In low flow hydrology, we ask successively two questions about a particular river site:

Seifu A. 22
Bahir Dar University, ioT School of Civil & Water Resources Engineering

• Does the river supply a particular water demand at all times?

• If not, how much water must be stored in order to meet any deficiency that may arise?

If the river is gauged, two methods can be employed for this task, flow duration analysis and Low flow
frequency analysis. For ungauged rivers, either data has to be transferred from nearby gauged rivers or
regional empirical equations need to be used.

3.3.1 Flow Duration Analysis

The flow duration curve is a plot that shows the percentage of time that flow in a stream is likely to
equal or exceed some specified value of interest. For example, it can be used to show the percentage of
time river flow can be expected to exceed a design flow of some specified value (e.g., 20 m3/s), or to
show the discharge of the stream that occurs or is exceeded some percent of the time (e.g., 80% of the
time).

The basic time unit used in preparing a flow-duration curve will greatly affect its appearance. For most
studies, mean daily discharges are used. These will give a steep curve. When the mean flow over a
long period is used (such as mean monthly flow), the resulting curve will be flatter due to averaging of
short-term peaks with intervening smaller flows during a month. Extreme values are averaged out
more and more, as the time period gets larger (e.g., for a flow duration curve based on annual flows at
a long-record station).

Step wise procedure of constructing FDCs

Step 1: Sort (rank) average daily discharges for period of record from the largest value to the smallest
value, involving a total of n values.
Step 2: Assign each discharge value a rank (M), starting with 1 for the largest daily discharge value.
Step 3: Calculate exceedence probability (P) as follows:

P = 100 * [ M / (n + 1) ]

P = the probability that a given flow will be equaled or exceeded (% of time)

M = the ranked position on the listing (dimensionless)
n = the number of events for period of record (dimensionless)

What does a FDC tell you about your stream?

A flow duration curve characterizes the ability of the basin to provide flows of various magnitudes.
Information concerning the relative amount of time that flows past a site are likely to equal or exceed a
specified value of interest is extremely useful for the design of structures on a stream. For example, a
structure can be designed to perform well within some range of flows, such as flows that occur
between 20 and 80% of the time (or some other selected interval).

The shape of the curve in the high-flow region indicates the type of flood regime the basin is likely to
have, whereas, the shape of the low-flow region characterizes the ability of the basin to sustain low
flows during dry seasons. A very steep curve (high flows for short periods) would be expected for
rain-caused floods on small watersheds. In the low-flow region, an intermittent stream would exhibit

Seifu A. 23
Bahir Dar University, ioT School of Civil & Water Resources Engineering

periods of no flow, whereas, a very flat curve indicates that moderate flows are sustained throughout
the year due to natural or artificial streamflow regulation, or due to a large groundwater capacity
which sustains the base flow to the stream.
Discharge, Q (m3/s)

Average flow

January December
Figure 3-2 Typical stream flow hydrograph

Adjusted flow duration curve due to pondage / Storage

Discharge, Q (m3/s)

Increased
Firm Yield
firm yield

0 25 50 100
Percent of time flow is equaled or exceeded
Figure 3-3 Flow duration curves
Example:
It is required to develop river “X” for small scale irrigation. Monthly flow duration curves have been
constructed for every month of the year using 25 years of data and the 75% dependable flows are
given in table below. ETo of the area is 4mm/day (take Kc =1). Effective rainfall over the irrigation
period (December to April) is nearly zero. Assuming an overall irrigation system efficiency of 40%
how many hectares of land can be irrigated at 75% dependability. The canal operates 12 hours per day.

Month Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Flow (m3/s) 0.12 0.10 0.08 0.07 0.08 0.29 1.30 2.07 1.32 0.49 0.26 0.17

Solution:
Let us find first the water requirement at the canal head.

Seifu A. 24
Bahir Dar University, ioT School of Civil & Water Resources Engineering

ETc = ETo = 4mm/day

Assuming 40% overall efficiency,
Water requirement at the canal head = 4/0.4 = 10mm/day
Discharge for 1ha, 1ha = 10000m2
= 10x10,000/(12x60x60)=2.315 l/s/ha
With the minimum design discharge of 0.07m3/s = 70l/s,
Command area = 70/2.315 = 30.24 ha

3.3.2 Low Flow Frequency Analysis

Reliability of a low-flow frequency estimate increases with the length of records. A record of 15 or 20
years may not provide a representative sample of low flows (Riggs 1985). In the absence of reliable
data at site of interest, regionalization approach is very useful to increase information for those sites
with inadequate data and much more for sites with no record. Frequency estimates can be done by
making use of theoretical distributions. In case of low flows,

P(X) = 1/T = Prob. (annual minimum < X) Eq. 3-5

The Extreme Value Type III distribution (also known as Gumbel’s Limited Distribution of the
Smallest Value) is used to generalize the regional information by low flow frequency growth factors
(curves). The EV3 distribution arises when the extreme is from a parent distribution that is limited in
the direction of interest. This distribution has found its greatest use in hydrology as the distribution of
low streamflows. Naturally low flows are bound by zero or other higher values on the left. The EV3
distribution can be expressed as:
XT = E + (U-E) [-ln(1-1/T)]1/A Eq. 3-6
where
XT = magnitude of a low flow to probability level of p,
The moment estimates of the parameters E (the lower limit), U and A (i.e., E, U and A) can be
obtained from the following equation:

U = Xm + Cv.Xm Y(a) Eq. 3-7

E = U – Cv.Xm.Z(a) , but E > 0
where
Xm = mean value of annual minimum flows,
Cv = coefficient of variation = standard deviation divided by the mean,

Y(a) , Z(a) and 1/A can be estimated from Table 4.1 provided by Haan (1977) as a function of the
sample skew coefficient. The same relation is depicted in Figure 3.4.

Alternatively, the parameters can be estimated using the following relationships:

Y(a) = (2-Cs)/6
Z(a) = [8+(3-Cs)3]/9 Eq. 3-8

Seifu A. 25
Bahir Dar University, ioT School of Civil & Water Resources Engineering

1/A = (1+Cs) / 3
where, Cs = skew coefficient calculated from the sample data

Table 3-1 Solutions of 1/a, Y(a) and Z(a) (Source: Haan, 1977)
Cs 1/a Y(a) Z(a) Cs 1/a Y(a) Z(a)
-0.638 0.10 0.425 8.737 1.160 0.70 0.142 1.549
-0.254 0.20 0.389 4.755 1.430 0.80 0.092 1.334
0.069 0.30 0.346 3.370 1.708 0.90 0.044 1.154
0.359 0.40 0.297 2.634 2.000 1.00 0.000 1.000
0.631 0.50 0.246 2.159 2.309 1.10 -0.040 0.867
0.896 0.60 0.193 1.815 2.640 1.20 -0.077 0.752

Parameters of EVIII Distribution

1.6 9.0
1.4 8.0
1.2 7.0
1.0 6.0
1/a, Y(a)

0.8 5.0 1/a

Z(a)
0.6 4.0 Y(a)
0.4 3.0 Z(a)
0.2 2.0
0.0 1.0
-0.2 0.0
-0.8 -0.4 0 0.4 0.8 1.2 1.6 2 2.4 2.8 3.2
Skew Coefficient (Cs)

Figure 3-4 Parameters of the EVIII distribution as a function of Cs

The value of a frequency analysis using the annual minimum series is increased by selecting
significant sequences of low flows, say 10 days, and taking the minimum 10-day average daily mean
flow in each year of the record.

Example:
The annual minimum daily mean flows for a river are given in table below. Compute the lean flows
having the following return periods: T = 5 years, T = 25 years

Annual
minimum 0.408 0.351 0.315 0.256 0.238 0.222 0.210 0.187 0.152 0.074
Flows
(m3/s)

Solution:
Compute the statistics of the flow flows:
Mean = Xm = 0.241

Seifu A. 26
Bahir Dar University, ioT School of Civil & Water Resources Engineering

St. Dev. = 0.098

Cv = 0.405
Cs = 0.114

Compute parameters of the EVIII distribution from Table, Graph or empirical relations.

Using the empirical relations as a function of Cs,

Y(a)= 0.314
Z (a) = 3.560
1/A = 0.371

U= 0.272
E= 0.000

Then, using the equation for XT of the EVIII distribution the following results are obtained;

T XT
5 0.156
10 0.118
15 0.101
20 0.090
25 0.083

3.4 Ungauged Catchments

In situations where dependable flows have to be estimated for ungauged catchments, estimations can
be made using either of the following:

• Transferring data from nearby gauged catchments of similar characteristics

• Empirical regional formulae
• Making lean flow measurements at the diversion site

Regionalization Method
In the absence of long-term historical data at the point of interest, it is common practice to resort to
other deterministic approaches and one of such method is the regionalization method. The basic
criteria in applying regionalization method are geographical and climatological similarity. It is
considered more realistic to rely on regionally available data rather than synthetically generated data
where its application is inevitable in the absence of data.

Based on the relationship given below the flow magnitude of un-gagged catchments can be estimated:

Seifu A. 27
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Aug
Qug= x Qg Eq. 3-9
AG
Where, Qg = Annual / Monthly flow of gauged site m3/s,
Qug= Annual flow of un-gagged river at proposed sites m3/s,
Ag = Area of the gauged catchment at the gauging station Km2,
Aug = Area of the un-gagged catchment at proposed sites Km2

Empirical Regional Formulae

The following regional empirical formulae have been derived by Admassu (1988) by studying a
number of gauged catchments:

For mean annual total flow:

Runoff = 0.85 X (area) 0.83 Eq. 3-10
Where,
Runoff is mean annual total flow (Mm3/year)

For low flows:

Eq. 3-11
Min30 = 2.08(area) 0.83
Min01 = 1.48 (area) 0.82
Where,
Min30 = 30 days average low flow in lit/sec
Min01 = 1 day average low flow in lit/sec

Making lean flow measurements at the diversion site

During the period of the study, lean flow measurements have to be carried out at the diversion site by
the hydraulic team. This is a very simple task and requires a float, a stop watch and a tape. This
measurement will be used to estimate the water availability or to verify water availability estimates by
the other two methods.
The following procedure should be used to apply floating method:
1. criteria for suitable place to estimate river discharge
a. straight stretch
b. far from any water control structures and bridge
c. uniform flow
d. wind velocity should be negligible at the time of measurement
2. measure the width of the stream at the selected site using a rope
3. divide the width into equal number of width say 30 or 40cm , and then measure the depth at
every width

Seifu A. 28
Bahir Dar University, ioT School of Civil & Water Resources Engineering

4. calculate the average depth; sum up the different depth and divide by the number of times
measurement is taken
5. cross-sectional area would be
a. width times average depth or
b. Find area of each section, since we divide the stream into small equal intervals. And
sum up to get the total area
6. From the site which is selected for width measurement, stretch a known length
7. The velocity of the stream is measured using the floating object. however, the velocity
measured by the floating object is surface velocity not the mean velocity
8. correction must be applied to the stream velocity measured in step 7 as follows
a. for depth less than 50cm, correction factor is 0.66
b. for depth between 50-60cm, use 0.70
c. for depth greater than 60cm, use 0.80
9. stream flow = average velocity*cross-sectional area

3.5 Demand-Water Supply Analysis

To design the intake of the irrigation scheme’s main canal, where it takes off from a diversion weir,
the discharge of the main canal will be needed. The discharge depends on types of crops and area of
the project. The procedure for determination of the off take canal capacity and hence the developed
area is based on demand-water supply analysis as follows:

A river lean discharge should be determined, as discussed in the previous sections.

The water right for the farmers downstream of the weir should be established from the
water and local authorities in the project area.

Determine the water requirement per hectare of the project area for the crop pattern advised
by the project agronomists.

Value of available water, which is lean flow determined in step one minus downstream
requirement in step 2, divided by value in step three gives the maximum area which can be
irrigated in a particular period.

Seifu A. 29
Bahir Dar University, ioT School of Civil & Water Resources Engineering

4 Design flood studies

4.1 Introduction
The study of design floods is very important in water resources engineering. Water supply structures, such
as, irrigation structures shall be designed to pass extreme flood events of given return periods. If the river
is gauged, with reasonably long records of flood data, say 20 to 30 years, then flood frequency analysis
could be used to estimate design floods.
In case of very short records or in the event that the river is not gauged, then hydrologists have to resort to
either of the following methods:

Transferring of data from nearby gauged catchments

Making use of flood marks and interviews

Rainfall – Runoff Methods
In this module, methods employed for the different cases will be discussed.

4.2 Flood Frequency Analysis

Flood frequency analysis is normally carried out using annual maximum series of flood peaks. The annual
maximum series takes the single maximum peak discharge in each year of record so that the number of
data values equals the record length in years.

4.2.1 Flood probabilities

When a series of annual maximum flows is subdivided by magnitude into discharge groups or classes of
class interval ∆Q, the number of occurrences of the peak flows (fi) in each class can be plotted against the
discharge values to give a frequency diagram. It is convenient to transform the ordinates of the diagram in
two ways: with respect both to the size of the discharge classes (∆Q) and the total number, N, of events in
the series. By plotting fi/(N. ∆Q) as ordinates, it is seen that the panel areas of the diagram will each be
given by fi/N, and hence the sum of those areas will be unity, i.e.,
n
fi N
∑ ( N .∆Q .∆Q)
i =1
=
N
=1 Eq. 4-1

Where n is the number of discharge classes.

If the data series is now imagined to be infinitely large in number and the class intervals are made
infinitesimally small, then in the limit, a smooth curve of the probability density distribution is obtained,
the area under the curve being unity, i.e.,
∞

∫ p(Q) dQ
0
=1 Eq. 4-2

Seifu A. 30
Bahir Dar University, ioT School of Civil & Water Resources Engineering

The probability that an annual maximum, Q, lies between two values, a and b, is given by:
∞

∫ p(Q) dQ
0
=1
Eq. 4-3
For any given magnitude, X, the probability that an annual maximum equals or exceeds X, i.e., that Q > X
is:
∞
P( X ) = ∫ p(Q) dQ
X Eq. 4-4
which is the area shaded under the probability curve.

If F(X) is the probability of Q < X:

X
F ( X ) = ∫ p (Q) dQ
0 Eq. 4-5,
and clearly P(X) = 1 – F(X) Eq. 4-6

P(X) is the probability of an annual maximum equaling or exceeding X in any given year, since it is the
relative proportion of the total number of annual maxima that have equaled or exceeded X. If X is
equaled or exceeded r times in N years (N large), then P(X) → r/N. The return period for X is, however,
T(X) = N/r. Thus:
P(X) = 1 / T(X) Eq. 4-7
It follows that
T (X) = 1 / P(X) = 1/ (1 – F(X)) Eq. 4-8

Seifu A. 31
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Thus, if T(X) = 100 years, P(X) = 0.01 and F(X) = 0.99.

4.2.2 Return Period and Risk

Flood frequency analysis entails the estimation of the peak discharge which is likely to be equaled or
exceeded on average once in a specified period, T years. This is called the T-year event and the peak, QT,
is said to have a return period or recurrence interval of T years. The return period, T years, is the long-
term average of the intervals between successive exceedances of a flood magnitude, QT, but it should be
emphasized that those intervals may vary considerably around the average value T. Thus a given record
may show 25-year events, Q25, occurring at intervals both much greater and much less than 25 years,
even in successive years.
Risk:
The probability R, called risk that Q > QT will occur at least once in n successive years, is given by:

R = 1 − (1 − 1/ T )n Eq. 4-9

Example:
What return period should an irrigation engineer adopt in the design of a culvert on a drain if he is willing
to accept only a 10% risk of flooding in the 25 years of expected life of the culvert?
Solution:
Using, R = 1 − (1 − 1/ T )n ; 0.10 = 1 − (1 − 1/ T )25
Solving for T gives, T = 238 years

4.2.3 Probability Distributions

A number of distribution functions have been suggested for the study of annual maximum (AM) flow
series. There is no clear cut guidance on the choice of distributions. A World Meteorological Organization
(WMO) survey of 1983 has shown that the EV1 distribution was the most commonly used distribution of
all followed by lognormal and LP3 with P3 a little less frequently used. In this module, the EVI and LP3
distributions, which are most commonly used, will be discussed
Table 4-1 Summary of frequency and extent of use of flood distributions. Source, WMO, 1989
EV1 EV2 EV3 GEV LN P3 Gamma 2/3 LP3
Agencies in which it is
used 28 11 8 7 27 17 6 22
Countries in which it is
used 16 9 8 2 16 12 6 13
Agencies in which it is
used as standard 18 3 0 5 16 9 3 17
Countries in which it is
used as standard in one or 10 3 0 1 8 7 3 7
more agencies

Seifu A. 32
Bahir Dar University, ioT School of Civil & Water Resources Engineering

4.2.4 Fitting theoretical distributions to sample data

Selected distributions can be fitted to sample data by either of the following methods:
• Graphical methods using special probability graph papers. Although subjective graphical curve
fitting by eye may be justifiably adequate in the analysis of inaccurately measured or even
estimated flood discharges.
• Analytical or objective methods of curve fitting, where the parameters of the selected distribution
are estimated from the sample statistic. The most commonly method is to relate parameters of the
distribution to the mean, standard deviation and skewness of the sample. This method is called the
Method of Moments.

4.2.5 The Graphical Approach

This method is illustrated by example taken from the book of Hydrology in Practice (Shaw, 1994). Table
below shows the Annual Maximum Series (X) of floods on a river in column 1. The values are put in
ascending order with the highest value having a rank of 1 and the lowest value a rank of N where N is the
number of the annual maxima. There are several formulae to compute the plotting positions, of which the
best is due to Gringorten given by:
P(X) = (r – 0.44)/ (N + 0.12) Eq. 5-10
Another formula which is more widely used is the one by Weibull given by:
P(X) = r / (N+1) Eq. 5-11
Both formulae have been applied in this example. The plotting positions are computed as shown in the
table.
The values are then plotted in special probability papers prepared for each distribution. In this case, the
EVI probability paper is used. The plot is done as shown in Fig.

Table 4-2 Annual Series and Plotting Positions

X (m3/s) Rank r P(X) F(X) P(X) F(X)

(Gringorten) (Weibull)
594.7 1 0.02 0.98 0.04 0.96
434.7 2 0.06 0.94 0.08 0.92
430.4 3 0.11 0.89 0.12 0.88
402.1 4 0.15 0.85 0.16 0.84
395.9 5 0.19 0.81 0.20 0.80
390.8 6 0.23 0.77 0.24 0.76
369.5 7 0.27 0.73 0.28 0.72
356.8 8 0.31 0.69 0.32 0.68
346.9 9 0.35 0.65 0.36 0.64

Seifu A. 33
Bahir Dar University, ioT School of Civil & Water Resources Engineering

342.9 10 0.40 0.60 0.40 0.60

342.6 11 0.44 0.56 0.44 0.56
321.4 12 0.48 0.52 0.48 0.52
318.0 13 0.52 0.48 0.52 0.48
317.2 14 0.56 0.44 0.56 0.44
300.2 15 0.60 0.40 0.60 0.40
290.0 16 0.65 0.35 0.64 0.36
284.6 17 0.69 0.31 0.68 0.32
283.2 18 0.73 0.27 0.72 0.28
273.3 19 0.77 0.23 0.76 0.24
256.3 20 0.81 0.19 0.80 0.20
196.8 21 0.85 0.15 0.84 0.16
190.3 22 0.89 0.11 0.88 0.12
185.5 23 0.94 0.06 0.92 0.08
138.8 24 0.98 0.02 0.96 0.04
Mean 323.5
St. Dev 97.1
N 24

Figure 4-1 Flood Frequency Gumbel Plot

Seifu A. 34
Bahir Dar University, ioT School of Civil & Water Resources Engineering

4.2.6 Computation of sample statistics:

The sample statistic that are computed in fitting distributions to sample data are the sample mean, sample
standard deviation and sample skewness:
Sample mean, X for a sample X of size N is computed as:
N
Xi
X =∑ Eq. 4-102
i =1 N

Sample Variance, s2 and sample standard deviation, s:

1 N 1 N
s2 = ∑ ( X i − X )2 and s = s =
N − 1 i =1
∑ ( X i − X )2
N − 1 i =1
Eq. 4-113

Sample Coefficient of variation (Cv) is a dimensionless dispersion parameters and is given by:
s
Cˆ v = Eq. 4-124
X
Sample Skew Coefficient,Cs ; A fully symmetrical distribution would exhibit the property that all odd
moments are equal to zero. A skewed distribution, however, would have an excessive weight to either side
of the center and the odd moments would exist. The third moment is thus used to define the skewness and
the skew coefficient.
N
N ∑ ( X i − X )3
Cs = i =1
Eq. 4-15
( N − 1) ( N − 2) s 3

4.2.7 Extreme Value Type I or Gumbel’s Distribution:

The equation of the EVI distribution is given by

−b( X −a )
F ( X ) = exp  −e 
 Eq. 4-16

where F(X) is the probability of an annual maximum Q < X as defined previously,

a, b are two parameters related to the moments of the population of Q values.

Defining the first moment (the mean) by µQ and the second moment (the variance) by σ2Q, the parameters
a and b are given by the following expressions:

Seifu A. 35
Bahir Dar University, ioT School of Civil & Water Resources Engineering

γ
a = µQ − (γ = 0.5772)
b
Eq. 4-17
π
b=
σQ 6

In the above equations, µQ and σ2Q pertain to the whole statistical population of floods at the station; with a
finite sample they can only be estimated from the moments of the data sample. Thus,
^ − 1 N
µQ = Q =
N
∑Q
i =1
i (the sample mean) Eq. 4-18

^
1 N −
σ 2 Q = sQ2 = ∑
N − 1 i =1
(Qi − Q ) 2 (the sample var iance) Eq. 4-19

Estimating the T – Year Flood

From the equation of the EVI distribution, the following relation can be derived:

1   T 
QT = a − ln ln  
b   T − 1  
Eq. 4-20

−
Substituting for the parameters a and b with the sample mean Q and standard deviation sQ as estimates of

the population values, then estimates of QT may be obtained from:

− 6    T  
QT = Q − sQ γ + ln ln     , which further simplifies to
π    T − 1  

− 6   T  
QT = Q + K (T ). sQ , where K (T ) = − γ + ln ln    Eq. 4-21
π    T − 1  

K(T) is called the frequency factor.

Table 4-3 The T – K(T) Relationship for Gumbel Distribution
T K(T) T K(T)
2 -0.16 50 2.61
5 0.72 100 3.14
10 1.30 200 3.68

Seifu A. 36
Bahir Dar University, ioT School of Civil & Water Resources Engineering

15 1.64 400 4.08

20 1.86 500
25 2.04 600 4.52
30 2.20 800 4.76
40 2.40 1000 4.94

Equation 4-21 is applicable to a sample of infinite size. When N is smaller, and is of a finite value, then
equation 4-21 is modified to vary K(T) as
T
− (γ + ln(ln( )))
K (T ) = T −1
π/ 6
Values of γ are given in the following table depending on N and its maximum value is 0.5772 as N
approaches infinity.

N 0 1 2 3 4 5 6 7 8 9
10 0.4952 0.4996 0.5035 0.5070 0.5100 0.5128 0.51570 0.51810 0.52020 0.55200
20 0.5236 0.5252 0.5268 0.5283 0.5296 0.5309 0.53200 0.53320 0.53430 0.53530
30 0.5362 0.5371 0.538 0.5388 0.5396 0.5402 0.54100 0.54180 0.54240 0.54300
40 0.5436 0.5442 0.5448 0.5453 0.5458 0.5463 0.54680 0.54730 0.54770 0.54810
50 0.5485 0.5489 0.5493 0.5497 0.5501 0.5504 0.55080 0.55110 0.55150 0.55180
60 0.5521 0.5524 0.5527 0.5530 0.5533 0.5535 0.55380 0.55400 0.55430 0.55450
70 0.5548 0.5550 0.5552 0.5555 0.5557 0.5559 0.55610 0.55630 0.55650 0.55670
80 0.5569 0.5570 0.5572 0.5574 0.5576 0.5578 0.55800 0.55810 0.55830 0.55850
90 0.5586 0.5587 0.5589 0.5591 0.5592 0.5593 0.55950 0.55960 0.55980 0.55990
100 0.5600

π / 6 is modified as shown in the following table depending on N.

N 0 1 2 3 4 5 6 7 8 9
10 0.9496 0.9676 0.9833 0.9971 1.0095 1.0206 1.0316 1.0411 1.0493 1.0565
20 1.0628 1.0696 1.0754 1.0811 1.0864 1.0915 1.0961 1.1004 1.1047 1.1086
30 1.1124 1.1159 1.1193 1.1226 1.1255 1.1285 1.1313 1.1339 1.1363 1.1388
40 1.1413 1.1436 1.1458 1.1480 1.1499 1.1519 1.1538 1.1557 1.1574 1.1590
50 1.1607 1.1623 1.1638 1.1658 1.1667 1.1681 1.1696 1.1708 1.1721 1.1734
60 1.1747 1.1759 1.1770 1.1782 1.1793 1.1803 1.1814 1.1824 1.1834 1.1844
70 1.1854 1.1863 1.1873 1.1881 1.1890 1.1898 1.1906 1.1915 1.1923 1.1930
80 1.1938 1.1945 1.1953 1.1959 1.1967 1.1973 1.1980 1.1987 1.1994 1.2001
90 1.2007 1.2013 1.2020 1.2026 1.2032 1.2038 1.2044 1.2049 1.2055 1.2060
100 1.2065

Seifu A. 37
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Example:
Assuming an EVI distribution where sample mean = 80 and sample standard deviation = 30 in m3/s :
a) What value of Q has exceedance probability 0.1?
b) What value of non-exceedance probability has Q = 140 m3/s?

Solution:
−
The relationship for EVI distribution is given by: QT = Q + K (T ). sQ
a) For exceedance probability, P = 0.1, T = 1/P = 1/0.1 = 10 years
From table above, K(T) = 1.3
Hence, Q10 = 80 + 1.3 * 30 = 119 m3/s

b) For Q = 140 m3/s, we want to find F (Q)

Parameters of the EVI distribution are computed as:

γ
a = µQ − (γ = 0.5772)
b π γ
b= = 0.0428 and a = µQ − = 80–0.5772/0.0428 =66.504
π σQ 6 b
b=
σQ 6

Hence,
− b( Q − a )
F (Q) = exp  −e  , substituting F(Q) = 0.958


4.2.8 Log-Pearson Type III Distribution

The Log-Pearson Type III distribution is another most commonly used statistical technique for fitting
frequency distribution data to predict the design flood for a river at some site.

The Log-Pearson Type III distribution is calculated using the general equation:

Log ( X T ) = Log ( X ) + Kσ Log ( x ) Eq. 4-22

where XT is the flood discharge value of some specified probability, is the average of the log x
discharge values, K is a frequency factor, and is the standard deviation of the log x values. The
frequency factor K is a function of the skewness coefficient and return period and can be found using the
frequency factor table. The flood magnitudes for the various return periods are found by solving the
general equation.

Seifu A. 38
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Example:
The annual daily maximum floods for Andassa River near Bahir Dar are given in Table below. You are
required to estimate design floods using EVI and LP3 distributions. Use the statistical method of moments
approach to estimate floods of 2, 5, 10, 50, 100, 200 return periods. Also estimate the design flood by
graphical method using EVI Probability paper.

Annual daily Maximum

Flood (m3/s)
Annual daily Maximum
Year Year Flood (m3/s)
1973 97 1990 160
1974 137 1991 154
1975 103 1992 213
1976 62
1993 114
1977 221 1994 108
1978 81 1995 102
1979 71 1996 176
1980 137
1997 74
1981 169 1998 156
1982 37 1999 116
1983 200 2000 331
1984 46
2001 90
1985 153 2002 52
1986 130 2003 152
1987 30
2004 111
1988 74
1989 144

Solution:
Annual Maximum Series
No. of Data 32

Annual daily Maximum Flood Plotting Position

Rank (m3/s) Ranked Flow (Gringorten) - F(X) Log (x)
1 97 331 0.983 2.520
2 137 221 0.951 2.344
3 103 213 0.920 2.328
4 62 200 0.889 2.301
5 221 176 0.858 2.246
6 81 169 0.827 2.228
7 71 160 0.796 2.204
8 137 156 0.765 2.193
9 169 154 0.733 2.188
10 37 153 0.702 2.185
11 200 152 0.671 2.182
12 46 144 0.640 2.158
13 153 137 0.609 2.137
14 130 137 0.578 2.137
15 30 130 0.547 2.114

Seifu A. 39
Bahir Dar University, ioT School of Civil & Water Resources Engineering

16 74 116 0.516 2.064

17 144 114 0.484 2.057
18 160 111 0.453 2.045
19 154 108 0.422 2.033
20 213 103 0.391 2.013
21 114 102 0.360 2.009
Annual daily Maximum Ranked Plotting Position
Rank Flood (m3/s) Flow (Gringorten) - F(X) Log (x)
22 108 97 0.329 1.987
23 102 90 0.298 1.954
24 176 81 0.267 1.908
25 74 74 0.235 1.869
26 156 74 0.204 1.869
27 116 71 0.173 1.851
28 331 62 0.142 1.792
29 90 52 0.111 1.716
30 52 46 0.080 1.663
31 152 37 0.049 1.568
32 111 30 0.017 1.477
Sample Statistics
Mean 125.03 2.04
St.
Dev. 62.44 0.23
CV. 0.50 0.11
CS -0.54

Step II: Estimation of floods

Analytical Method

I: EV I distribution using statistics of untransformed series

T K(T) Q (T)
2 -0.16 115.04
5 0.72 169.98
10 1.30 206.20
25 2.04 252.40
50 2.61 287.99
100 3.14 321.08
200 3.68 354.79

II : LP 3 distribution – using statistics of log-transformed series; Cs ≈ - 0.50

T K(T) Log (QT) Q(T)

2 0.083 2.061 115.14
5 0.856 2.241 174.14

Seifu A. 40
Bahir Dar University, ioT School of Civil & Water Resources Engineering

10 1.216 2.325 211.15

25 1.567 2.406 254.79
50 1.777 2.455 285.10
100 1.955 2.496 313.60
200 2.108 2.532 340.36
Graphical Method: Plot the ranked discharge-versus plotting position on EVI plot and fit a straight line
by eye-fitting. Then the flood quantiles for various return periods can be read from the curve. This is left
as an exercise to the trainee.

However, it can be concluded that the analytic approach is recommended as it is not subject to personal
judgment as in the case of manually fitting the curve in the graphical approach.

4.2.9 Instantaneous versus Daily Peaks

In most cases the peak flow data that is recorded and available are mean daily maximum flows. After
making the frequency analysis, the mean daily flood peak of a given return period has to be converted to
Instantaneous or Momentary Peaks. According to studies made in projects in the Abay and Tekeze basins,
the ratio between instantaneous peaks and daily peaks varied between 1.13 and 1.48 with an average value
of 1.24. Hence, this average value can be used as a peaking factor.

Hence, in the previous example the instantaneous peaks could be computed by using a multiplication
factor of 1.24.

Maximum Daily Peaks Instantaneous Peaks

T EVI LPIII EVI LPIII
2 115.04 115.14 142.6 142.77
5 169.98 174.14 210.8 215.93
10 206.20 211.15 255.7 261.82
25 252.40 254.79 313.0 315.94
50 287.99 285.10 357.1 353.52
100 321.08 313.60 398.1 388.86
200 354.79 340.36 439.9 422.04

Seifu A. 41
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Frequency Factors K for Log-Pearson Type III Distribution (Haan, 1977, Table 7.7)
Recurrence Interval In Years
1.0101 2 5 10 25 50 100 200
SKEW Percent Chance (>=) = 1-F
COEFFICIENT
Cs 99 50 20 10 4 2 1 0.5
3 -0.667 -0.396 0.420 1.180 2.278 3.152 4.051 4.970
2.9 -0.690 -0.390 0.440 1.195 2.277 3.134 4.013 4.904
2.8 -0.714 -0.384 0.460 1.210 2.275 3.114 3.973 4.847
2.7 -0.740 -0.376 0.479 1.224 2.272 3.093 3.932 4.783
2.6 -0.769 -0.368 0.499 1.238 2.267 3.071 3.889 4.718
2.5 -0.799 -0.360 0.518 1.250 2.262 3.048 3.845 4.652
2.4 -0.832 -0.351 0.537 1.262 2.256 3.023 3.800 4.584
2.3 -0.867 -0.341 0.555 1.274 2.248 2.997 3.753 4.515
2.2 -0.905 -0.330 0.574 1.284 2.240 2.970 3.705 4.444
2.1 -0.946 -0.319 0.592 1.294 2.230 2.942 3.656 4.372
2 -0.990 -0.307 0.609 1.302 2.219 2.912 3.605 4.298
1.9 -1.037 -0.294 0.627 1.310 2.207 2.881 3.553 4.223
1.8 -1.087 -0.282 0.643 1.318 2.193 2.848 3.499 4.147
1.7 -1.140 -0.268 0.660 1.324 2.179 2.815 3.444 4.069
1.6 -1.197 -0.254 0.675 1.329 2.163 2.780 3.388 3.990
1.5 -1.256 -0.240 0.690 1.333 2.146 2.743 3.330 3.910
1.4 -1.318 -0.225 0.705 1.337 2.128 2.706 3.271 3.828
1.3 -1.383 -0.210 0.719 1.339 2.108 2.666 3.211 3.745
1.2 -1.449 -0.195 0.732 1.340 2.087 2.626 3.149 3.661
1.1 -1.518 -0.180 0.745 1.341 2.066 2.585 3.087 3.575
1 -1.588 -0.164 0.758 1.340 2.043 2.542 3.022 3.489
0.9 -1.660 -0.148 0.769 1.339 2.018 2.498 2.957 3.401
0.8 -1.733 -0.132 0.780 1.336 1.993 2.453 2.891 3.312
0.7 -1.806 -0.116 0.790 1.333 1.967 2.407 2.824 3.223
0.6 -1.880 -0.099 0.800 1.328 1.939 2.359 2.755 3.132
0.5 -1.955 -0.083 0.808 1.323 1.910 2.311 2.686 3.041
0.4 -2.029 -0.066 0.816 1.317 1.880 2.261 2.615 2.949
0.3 -2.104 -0.050 0.824 1.309 1.849 2.211 2.544 2.856
0.2 -2.178 -0.033 0.830 1.301 1.818 2.159 2.472 2.763
0.1 -2.252 -0.017 0.836 1.292 1.785 2.107 2.400 2.67
0 -2.326 0.000 0.842 1.282 1.751 2.054 2.326 2.576

Seifu A. 42
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Frequency Factors K for Log-Pearson Type III Distribution (Haan, 1977, Table 7.7)
Recurrence Interval In Years
1.0101 2 5 10 25 50 100 200
SKEW Percent Chance (>=) = 1-F
COEFFICIENT
Cs 99 50 20 10 4 2 1 0.5
-0.1 -2.400 0.017 0.846 1.27 1.716 2.000 2.252 2.482
-0.2 -2.472 0.033 0.850 1.258 1.680 1.945 2.178 2.388
-0.3 -2.544 0.050 0.853 1.245 1.643 1.890 2.104 2.294
-0.4 -2.615 0.066 0.855 1.231 1.606 1.834 2.029 2.201
-0.5 -2.686 0.083 0.856 1.216 1.567 1.777 1.955 2.108
-0.6 -2.755 0.099 0.857 1.200 1.528 1.720 1.880 2.016
-0.7 -2.824 0.116 0.857 1.183 1.488 1.663 1.806 1.926
-0.8 -2.891 0.132 0.856 1.166 1.448 1.606 1.733 1.837
-0.9 -2.957 0.148 0.854 1.147 1.407 1.549 1.660 1.749
-1 -3.022 0.164 0.852 1.128 1.366 1.492 1.588 1.664
-1.1 -3.087 0.180 0.848 1.107 1.324 1.435 1.518 1.581
-1.2 -3.149 0.195 0.844 1.086 1.282 1.379 1.449 1.501
-1.3 -3.211 0.210 0.838 1.064 1.240 1.324 1.383 1.424
-1.4 -3.271 0.225 0.832 1.041 1.198 1.270 1.318 1.351
-1.5 -3.33 0.240 0.825 1.018 1.157 1.217 1.256 1.282
-1.6 -3.880 0.254 0.817 0.994 1.116 1.166 1.197 1.216
-1.7 -3.444 0.268 0.808 0.970 1.075 1.116 1.140 1.155
-1.8 -3.499 0.282 0.799 0.945 1.035 1.069 1.087 1.097
-1.9 -3.553 0.294 0.788 0.920 0.996 1.023 1.037 1.044
-2 -3.605 0.307 0.777 0.895 0.959 0.980 0.990 0.995
-2.1 -3.656 0.319 0.765 0.869 0.923 0.939 0.946 0.949
-2.2 -3.705 0.330 0.752 0.844 0.888 0.900 0.905 0.907
-2.3 -3.753 0.341 0.739 0.819 0.855 0.864 0.867 0.869
-2.4 -3.800 0.351 0.725 0.795 0.823 0.830 0.832 0.833
-2.5 -3.845 0.360 0.711 0.711 0.793 0.798 0.799 0.800
-2.6 -3.899 0.368 0.696 0.747 0.764 0.768 0.769 0.769
-2.7 -3.932 0.376 0.681 0.724 0.738 0.740 0.740 0.741
-2.8 -3.973 0.384 0.666 0.702 0.712 0.714 0.714 0.714
-2.9 -4.013 0.390 0.651 0.681 0.683 0.689 0.690 0.690
-3 -4.051 0.396 0.636 0.660 0.666 0.666 0.667 0.667

Seifu A. 43
Bahir Dar University, ioT School of Civil & Water Resources Engineering

4.3 Un-gauged Catchments

In many situations, the hydraulic engineer/hydrologist has to carry out river development studies in un-
gauged streams. It is said that “Engineering is the solution of practical problems with insufficient data.”
Hence, the engineer has to find some means of estimating the required floods. Some approaches that could
be used in such situations will be discussed in this module.

4.3.1 Regional flood frequency estimation

The method of regional flood frequency analysis has been used to estimate design floods at ungauged sites
or at sites with only a few years of data. Recently, regional estimation is being used to pool regional data
to give QT estimates which have low standard errors than those obtainable from at-site data alone.

Index flood procedure:

The variate Q is divided by Qbar and the resulting variate X = Q/Qbar is assumed to have the same form
of distribution at every site. In this context, Qbar is the index flood. Then,

QT = Qbar . XT Eq. 5-23

Where
QT is flood magnitude of given return period at any site
Qbar is the mean annual daily maximum flood at the site
XT is estimated from the pooled data

For ungauged catchment the problem reduces to estimating the mean annual maximum daily flood. This
can be estimated by relating the catchment characteristics (area, slope) with the gauged catchments.

One such empirical formula is the one derived by Admasu (1988), which gave an empirical relationship
for Qbar by analyzing a number of catchments in different river basins of Ethiopia.

This formula is given by:

Qbar = 0.87 * (Area) 0.70 , Eq. 5-24, where A is catchment area in km2, and Qbar is in m3/s.

4.3.2 Flood marks and interviews

The mode of the Gumbel annual flood frequency distribution (Return period ~ 1.5 years) generally lies at
bank full level, the elevation of the natural levee, or the natural bank height. This knowledge allows fixing
one point of the frequency distribution on the Gumbel probability paper.

Seifu A. 44
Bahir Dar University, ioT School of Civil & Water Resources Engineering

The second point should be derived from interviews in the field. Detailed questioning of as many
inhabitants of the area as possible may give you an indication of the largest flood in say 20 years.
Interviews should contain questions like:

• In which year do you remember the biggest flood?

• What age were you then?
• What age are you now?
• Where were you at the time of the flood?
• Can you indicate the level of the flood (on a tree, a house, a wall)?
Care should be taken to interview the right people.

Such flood marks should also be observed in the river reach both upstream and downstream of the site.
Once the two points are indicated, a straight line can be drawn and extrapolated to estimate design floods
of the required return periods.

4.3.3 Using Regional Formulae

Some regional studies might have been done in the study area to derive flood magnitude relationships of
gauged catchments as a function of catchment characteristics. In the simplest for the relations would be a
function of catchment area. This type of relationships could also be used with caution. For example, in the
Master Plan Study of The Abay Basin (BCEOM, 1999) the flood magnitudes were given by the
relationship QT = a * (Area) b, where Area is the catchment area, a and b are regression coefficients as
shown in Table below.

Table 4-4 Regional Analysis of flood discharge for the Abay Basin (BCEOM, 1999)
t (years) a B
Catchment area < 10,000 km2 100 47.8 0.339
1000 75.7 0.313
10000 105.0 0.296

4.3.4 Rainfall Runoff Methods

Another most commonly used method is to estimate design floods from rainfall. Design rainfall of the
required return period is estimated using frequency analysis techniques described in the previous sections
and deterministic methods are used to derive the floods from the rainfall. This method is commonly
known as rainfall-runoff method.

Two such methods will be discussed in this module – the Rational Method and the Soil Conservation
Service (SCS) method.

Seifu A. 45
Bahir Dar University, ioT School of Civil & Water Resources Engineering

4.3.4.1 The Rational Method

The Rational Method can be applied to small catchments not greater than 5 km2. The consequences of
applying the Rational Method to larger catchments is to produce an over estimate of discharge and a
conservative design. The method is nevertheless frequently used in standard or modified form for much
larger catchments. This is because of its relative simplicity. The vast majority of catchments producing
floods imposed on the farm drainage system lie within the validity of the Rational Method and it has been
used as the principal method of estimating design discharges of dykes, culverts, drainage channels, etc.

The Rational Method is based on the following formula:

QT = 0.2778 C.RT.A.Fr Eq. 5-25

where
QT = peak flow corresponding to return period of T years in m3/sec;
C = a 'runoff' coefficient expressing the ratio of rate of runoff to rate of rainfall;
RT = maximum intensity of rainfall in mm/hr, for a duration equal to the time of concentration;
A = drainage area in km2;
Fr = is the areal reduction factor (this factors improves the catchment limitations imposed on
the use of rational method).
Coefficient of runoff C for the formula are given by many soil and water conservation texts.
Table 4-5 Runoff coefficient, C, for clay and silt loam soil watershed

Class Topography and vegetation Runoff

percentage
Wood land

Flat (0-5%) 30

Rolling (5-10%) 35

Hilly (10-30%) 50
Pasture

Flat 30

Rolling 36

Hilly 42
Cultivated

Flat 50

Rolling 60

Hilly 72

The time of concentration can be estimated by the following formula:

Tc = 0.00025x(L/S0.5)0.80 (Kirpich / Ramser formula) Eq. 5-26

Where,

Seifu A. 46
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Tc = time of concentration (in hours),

L = maximum length of main stream (in meters),
S = the overall catchment slope (in m/m).

Maximum rainfall magnitudes of 24 hr duration corresponding to various return periods (or probability
levels) can be estimated by applying the following form:

RT = Rm [1+ CV.K(T)] Eq. 5-27

where
RT = rainfall intensity corresponding to return period of T years (mm),
Rm = mean of annual maximum intensity of rainfall (mm/hr),
CV = coefficient of variation of annual intensity of rainfall,
K(T) = frequency factor

For Extreme Value Type I (or Gumbel) Distribution,

K(T) = -0.779 {0.577 + ln.ln[T/{(T –1}]} Eq. 5-28

The rainfall intensities corresponding to duration D hours can be estimated by empirically derived formula
derived from continuous records of rainfall. One such formula is expressed as:

Ih = I24.(4.89/D-0.5 ) Eq. 5-29

where
Ih = maximum rainfall intensity corresponding to duration D (mm/hr),
D = duration of Ih (hr),
I24 = maximum rainfall intensity corresponding to duration 24 hours (mm/hr)

Maximum rainfall intensity of the required duration can also be estimated from Intensity – Duration –
Frequency Curves. The following IDF Table developed by the Ethiopian Roads Authority can also be
used.

The area reduction factor (Fr) is introduced to account for the spatial variability of point rainfall over the
catchment. This is not significant for small catchments but becomes so as catchment size increases. The
relationship adopted for ‘Fr’ is based on that developed for the East African condition (Fiddes, 1997). The
relationship can be expressed as:

Fr = 1 - 0.02 D-0.33 A 0.50 Eq. 5-30

where
D = duration in hours;

Seifu A. 47
Bahir Dar University, ioT School of Civil & Water Resources Engineering

A = drainage area in km2

This equation applies for storms of up to 8 hours duration. For longer durations on large catchments the
value of D can be taken as 8 for use in the above formula.
Table 4-6 Rainfall Intensity- Duration- Frequency for Ethiopia (Ethiopian Roads Authority)
Duration Return Period (Years) / Rainfall Intensity (mm/hr)
(Minutes) 5 10 25 50 100
0 125.3 162.7 200.0 226.0 250.0
5 125.3 162.7 200.0 226.0 250.0
10 110.7 137.3 162.7 183.9 203.4
20 88.0 108.0 125.3 141.6 156.6
30 70.0 86.7 102.7 116.1 128.4
40 57.3 72.7 86.7 98.0 108.4
50 47.7 62.0 75.3 85.1 94.1
60 40.7 53.3 66.7 75.4 83.4
70 35.3 47.3 60.7 68.6 75.9
80 32.0 43.3 56.0 63.3 70.0
90 29.0 40.0 51.0 57.6 63.8
100 27.3 37.0 48.0 54.2 60.0
110 25.0 34.7 44.7 50.5 55.9
120 23.3 32.0 42.7 48.3 53.4
Notes: I (50) = 1.13 * I (25)
I (100) = 1.25 * I (25)

4.3.4.2 The SCS Method

This method is more physically based than the Rational Method and has wider application. The step by
step procedure of this method is presented below.

Step I – Estimation of Design Storm

The design storm is characterized by its duration and temporal distribution. In the absence of continuous
records of rainfall, the design storm can be derived from the 24 hour rainfall. According to the SCS, four
distributions are developed for the entire United States. Other countries have to determine which of the
four will best fit the local condition or develop their own relationships. For Ethiopia, use a graph shown
on page 13.

The duration of the design storm is given in terms of the time of concentration as,
D = Tc / 7.5, Tc can be determined using the Kirpich formula.

Seifu A. 48
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Figure 4-2 SCS Rainfall Distributions

Figure 4-3: A typical Rainfall Hyetograph of 30 minutes duration

Step II – Estimation of Excess Rainfall

According to this method, the cumulative excess rainfall (the volume of direct runoff) resulting from a
rainfall (cumulative) of P is given by:

( P − 0.2 S )2
R= for P > 0.2 S , or R = 0, otherwise. Eq. 5-31
P + 0.8S
where, S is the potential retention in the basin.
The potential retention S is expressed in terms of a dimensionless curve number CN, which depends on
soil type and land use/cover in the drainage basin. The relation between S and CN is defined, in metric
units as:
25, 400
S= − 254 ( when water depths are exp ressed in mm) Eq. 5-32
CN

Seifu A. 49
Bahir Dar University, ioT School of Civil & Water Resources Engineering

The curve number is a function of the hydrologic soil group and antecedent moisture conditions of the
catchment. The following table shows the curve number values for the four hydrological soil groups.
Three classes of antecedent moisture condition (AMC) are defined, dry (AMC I), average (AMC II), and
wet (AMC III).

Group A Soils: High infiltration (low runoff). Sand, loamy sand, or sandy loam. Infiltration rate > 0.3
inch/hr when wet.
Group B Soils: Moderate infiltration (moderate runoff). Silt loam or loam. Infiltration rate 0.15 to 0.3
inch/hr when wet.
Group C Soils: Low infiltration (moderate to high runoff). Sandy clay loam. Infiltration rate 0.05 to
0.15 inch/hr when wet.
Group D Soils: Very low infiltration (high runoff). Clay loam, silty clay loam, sandy clay, silty clay, or
clay. Infiltration rate 0 to 0.05 inch/hr when wet.
Table 4-7: Runoff Curve Numbers (SCS, 1986) For AMC II
Description of Land Use Hydrologic Soil Group
A B C D
Paved parking lots, roofs, driveways 98 98 98 98
Streets and Roads:
Paved with curbs and storm sewers 98 98 98 98
Gravel 76 85 89 91
Dirt 72 82 87 89
Cultivated (Agricultural Crop) Land*:
Without conservation treatment (no terraces) 72 81 88 91
With conservation treatment (terraces, contours) 62 71 78 81
Pasture or Range Land:
Poor (<50% ground cover or heavily grazed) 68 79 86 89
Good (50-75% ground cover; not heavily grazed) 39 61 74 80
Meadow (grass, no grazing, mowed for hay) 30 58 71 78
Brush (good, >75% ground cover) 30 48 65 73
Woods and Forests:
Poor (small trees/brush destroyed by over-grazing or burning) 45 66 77 83
Fair (grazing but not burned; some brush) 36 60 73 79
Good (no grazing; brush covers ground) 30 55 70 77
For the other two antecedent conditions, i.e., AMC I and AMC III, the following conversion table can be
used.

Seifu A. 50
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Table 4-8: Runoff Curve Numbers for AMC I and AMC III
Corresponding CNS Corresponding CNS
CN for AMC II AMC I AMC III CN for AMC II AMC I AMC III
100 100 100 50 31 70
95 87 98 45 26 65
90 78 96 40 22 60
85 70 94 35 18 55
80 63 91 30 15 50
75 57 88 25 12 43
70 51 85 20 9 37
65 45 82 15 6 30
60 40 78 10 4 22
55 35 74 5 2 13

Step III – Developing the SCS – Dimensionless Unit Hydrograph

A unit hydrograph represents the runoff resulting from:
* One cm of precipitation excess,
* Generated uniformly over the watershed,
* At a uniform rate,
* For a duration D.

The hydrograph is made dimensionless by expressing:

* Ordinates as a fraction of the peak discharge qp,
* Time axis as a fraction of the time-to-peak Tp.
By analyzing a large amount of measured data the SCS developed an average dimensionless unit
hydrograph:
t/tp Q/Qp t/tp Q/Qp
0 0 1.3 0.84
0.1 0.015 1.4 0.75
0.2 0.075 1.5 0.66
0.3 0.16 1.6 0.56
0.4 0.28 1.8 0.42
0.5 0.43 2 0.32
0.6 0.6 2.4 0.18
0.7 0.77 3 0.075
0.8 0.89 3.5 0.036
0.9 0.97 4 0.018
1 1 4.5 0.009
1.1 0.98 5 0.004
1.2 0.92
Figure 4-4 The SCS – Dimensionless Unit Hydrograph
To dimension the time axis of the UH we use the following relationships:

Tp = 5 D and Tp = 2/3 Tc therefore D = Tc / 7.5 Eq. 5-33

Seifu A. 51
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Tp=Time to peak [hours]

Tc=Time of concentration [hours]
D=Burst duration [hours]

The peak discharge qp of the unit hydrograph is given by:

Q p = 2 .0 8 A / t p Eq. 5-34

where Qp is discharge in m3/s/cm

A is catchment area in km2
Tp is time to peak in hours

Step IV – Convolution of the excess rainfall hyetograph using the Unit Hydrograph
The convolution equation is given by:

n≤M
Qn = ∑PU
m =1
m n − m +1 Eq. 5-35

A more illustrative approach of the convolution procedure is shown in Figure 4.5 below.

Figure 4-5 UH Convolution procedure

Example

Seifu A. 52
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Near the town of Quiha a reservoir is planned to be built. It is necessary to get the flood discharge for 100-
year return period of for the design of the spillway. Using the complex hydrograph, determine the peak
discharge for the watershed of 30km2 with CN = 92.5.

Solution

Maximum Design Rainfall

24-hr rainfall for 100 year return period is 95mm at Quiha.

Time of concentration
The longest water-course would be divided into several intervals and time of concentration coputed for
each and the total time using Kirpich formula described in Table 2-4. The time of concentration is 3hr.

Design Rainfall Arrangement

The maximum rainfall increments would be one hour and they would be arranged as follows as described
in section 2-2-3:

Time Accumulated Rainfall Incremental Descending Rearranged Rearranged Rainfall

(hr) (mm) Rainfall order order
% (mm) (mm) Incremental Cumulative
0-1 35 32.2 32.2 1 6 4.6 4.6
1-2 48 44.2 12.0 2 4 6.4 11.0
2-3 58 53.4 9.2 3 3 9.2 20.2
3-4 65 59.8 6.4 4 1 32.2 52.4
4-5 70 64.4 4.6 5 2 12.0 64.4
5-6 75 69.0 4.6 6 5 4.6 69.0
6-12 90 82.8 13.8
12-24 100 92.0 9.2

Estimation of Direct runoff

Using equation 4-31, direct runoff is estimated as follows:

Time (hr) Incremental Cumulative Cumulative Direct Incremental Direct

Rainfall (mm) rainfall (mm) Runoff (mm) runoff (mm)
0-1 4.6 4.6 0.01 0.01
1-2 6.4 11.0 1.72 1.71

Seifu A. 53
Bahir Dar University, ioT School of Civil & Water Resources Engineering

2-3 9.2 20.2 7.05 5.33

3-4 32.2 52.4 33.84 26.79
4-5 12.0 64.4 42.84 9.00
5-6 4.6 69.0 44.44 1.60

Computation of hydrograph for each increment of runoff

It is based on the dimensionless SCS unit hydrograph method. For each time interval, D, compute the time
of peak, Tp, base time Tb, and the peak discharge qp, for 1mm of runoff.

D 1 .0
TP = + 0.6 * TC = + 0.6 * 2.7 = 2.12hours
2 2

Tb = 2.67TP = 2.67 * 2.12 = 5.66hours

0.208 A 0.208 * 30
qP = = = 2.97 m3 / s.mm
TP 2.12

Peaks for each incremental direct runoff are obtained by multiplying the peak discharge for 1mm of runoff
computed above by the increment.

Time (hr) Incremental qP for 1mm qP for Begin time Peak time End time
Runoff runoff incremental
(mm) (m3/s) runoff (hr) (hr) (hr)
(m3/s)
0-1 0.01 2.97 0.03 0 2.12 5.66
1-2 1.71 2.97 5.08 1 3.12 6.66
2-3 5.33 2.97 15.83 2 4.12 7.66
3-4 26.79 2.97 79.56 3 5.12 8.66
4-5 9.00 2.97 26.73 4 6.12 9.66
5-6 1.60 2.97 4.75 5 7.12 10.66

Seifu A. 54
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Design Hydrograph

Seifu A. 55
Bahir Dar University, ioT School of Civil & Water Resources Engineering

5 Flow Routing

One of the most common problems facing a practicing civil engineer is the estimation of the hydrograph
of the rise and fall of a river at any given point on the river during the course of a flood event. The
problem is solved by the techniques of flood routing.

Flood Routing is a process used to predict the temporal and spatial variations of a flood hydrograph as it
moves through a river reach or reservoir. A flood hydrograph is modified in two ways as the storm water
flows downstream. Firstly, the time of the peak rate of flow occurs later at downstream points. This is
known as translation. Secondly, the magnitude of the peak rate of flow is diminished at downstream
points, the shape of the hydrograph flattens out, and the volume of flood waters takes longer to pass a
lower section. This modification to the hydrograph is called attenuation.

Figure 5-1:Discharge hydrograph routing effects

5.1 Methods of flood routing:

In general, routing techniques may be classified into two categories: hydraulic or distributed routing,
and hydrologic or lumped routing. Hydraulic routing techniques are based on the solution of the partial
differential equations of unsteady open channel flow. These equations are often referred to as the St.
Venant equations or the dynamic wave equations. Hydrologic routing employs the continuity equation and
an analytical or an empirical relationship between storage within the reach and discharge at the outlet.

Seifu A. 56
Bahir Dar University, ioT School of Civil & Water Resources Engineering

5.2 Applications of flood routing:

The design engineer needs to be able to route flood hydrographs:

To forecast floods in lower parts of a river basin

To assess the capacity of reservoir spillways

In designing flood protection schemes or in evaluating the span and height of bridges or other river
structures

For such purposes, the upstream hydrograph is routed through the reach with a hydrologic routing
technique that predicts changes in hydrograph shape and timing. Local flows are then added at the
downstream location to obtain the total flow hydrograph. This type of approach is adequate as long as
there are no significant backwater effects or discontinuities in the water surface because of jumps or bores.
When there are downstream controls that will have an effect on the routing process through an upstream
reach, the channel configuration should be treated as one continuous system. This can only be
accomplished with a hydraulic routing technique that can incorporate backwater effects as well as internal
boundary conditions, such as those associated with culverts, bridges, and weirs.

In this module, only hydrologic routing techniques will be discussed.

5.3 Hydrologic Routing Techniques

Hydrologic routing employs the use of the continuity equation and either an analytical or an empirical
relationship between storage within the reach and discharge at the outlet. In its simplest form, the
continuity equation can be written as inflow minus outflow equals the rate of change of storage within the
reach:
∆S
I −O= Eq. 5-1
∆t
where
I = the average inflow to the reach during ∆t
O = the average outflow from the reach during ∆t
S = storage within the reach

5.3.1 Modified puls Reservoir Routing

One of the simplest routing applications is the analysis of a floodwave that passes through an unregulated
reservoir. The inflow hydrograph is known, and it is desired to compute the outflow hydrograph from the

Seifu A. 57
Bahir Dar University, ioT School of Civil & Water Resources Engineering

reservoir. Assuming that all gate and spillway openings are fixed, a unique relationship between storage
and outflow can be developed, as shown in Figure 5.2.

The equation defining storage routing, based on the principle of conservation of mass, can be written in
approximate form for a routing interval t. Assuming the subscripts “1” and “2” denote the beginning and
end of the routing interval, the equation is written as follows:
O1 + O2 I1 + I 2 S 2 − S1
− = Eq. 5-2
2 2 ∆t

Figure 5-2 Reservoir storage routing

The known values in Equation 5.2 are the inflow hydrograph and the storage and discharge at the
beginning of the routing interval. The unknown values are the storage and discharge at the end of the
routing interval. With two unknowns (O2 and S2) remaining, another relationship is required to obtain a
solution. The storage-outflow relationship is normally used as the second equation. How that relationship
is derived is what distinguishes various storage routing methods.

The modified puls method applied to reservoirs consists of a repetitive solution of the continuity equation.
It is assumed that the reservoir water surface remains horizontal, and therefore, outflow is a unique
function of reservoir storage. The continuity equation, Equation 5-2, can be manipulated to get both of the
unknown variables on the left-hand side of the equation:

 S 2 O2   S1 O1  I1 + I 2
 + = + + − O1 Eq. 5-3
 ∆t 2   ∆t 2  2

S O  I1 + I 2
Replacing  +  by G and by Im Equation 5-3 can be written:
 ∆t 2  2

G2 = G1 + I m − O1 Eq. 5-4

Seifu A. 58
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Since I is known for all time steps, and O1 and S1 are known for the first time step, the right-hand side of
the equation can be calculated. The left-hand side of the equation can be solved by trial and error. This is
accomplished by assuming a value for either S2 or O2, obtaining the corresponding value from the storage-
outflow relationship, and then iterating until Equation 5-3 or 5-4 is satisfied.

Figure 5-3 Reservoir storage – outflow curve

The stepwise procedure for applying the modified puls method to reservoirs can be summarized as follows:
(a) Determine a composite discharge rating curve for all of the reservoir outlet structures.
(b) Determine the reservoir storage that corresponds with each elevation on the rating curve for reservoir outflow.
(c) Select a time step and construct a storage-indication versus outflow curve [(S/∆ t) + (O/2)] versus O.
(d) Route the inflow hydrograph through the reservoir based on Equation 5-3 and the storage-indication curve.

O2

O1

G1 G2 = G1+Im – O1

S/∆t + O/2 = G

Figure 5-4 Storage Indication vs outflow curve (Auxiliary Curve)

The method will be illustrated by the following example.

Example – Reservoir Routing

A reservoir has the elevation-area-storage relationship above the spillway crest as shown in Table 1
below. The spillway crest is B = 70m wide and the discharge over the spillway is given by Q = C B H1.5 ,

Seifu A. 59
Bahir Dar University, ioT School of Civil & Water Resources Engineering

where H is head over the spillway crest and C is discharge coefficient which can be taken as 2.14 for all
heads. For the inflow hydrograph given in Table 2, find the outflow hydrograph.

Table 1 Table 2
Elevation above Area Storage Storage
Time Inflow
spillway crest (Sq.Km) (Cu.m)) (MCM)
(h) (m3/s)
0 3.6 0 0
0 12
0.1 3.65 362500 0.36
1 42
0.2 3.7 730000 0.73
2 72
0.3 3.75 1102500 1.10
3 102
0.4 3.8 1480000 1.48
4 87
0.5 3.85 1862500 1.86
5 72
0.6 3.9 2250000 2.25
6 57
0.7 3.95 2642500 2.64
7 42
0.8 4 3040000 3.04
8 27
0.9 4.05 3442500 3.44
9 12
1 4.1 3850000 3.85
10 12
1.1 4.15 4262500 4.26
1.2 4.2 4680000 4.68
1.4 4.3 5530000 5.53

Solution:

Step I : Select time step, in this case 1 hr = 3600 s is selected.

Step II: Prepare the Auxiliary Curve
h S O G = S/dt + O/2 Auxilary Curve
0 0 0 0 160
0.1 362500 4.74 103.06 140
0.2 730000 13.40 209.48 120
Outflow, O

0.3 1102500 24.61 318.56 100

0.4 1480000 37.90 430.06 80
60
0.5 1862500 52.96 543.84
40
0.6 2250000 69.62 659.81 20
0.7 2642500 87.73 777.89 0
0.8 3040000 107.19 898.04 0 100 200 300 400 500 600 700 800 900 1000
0.9 3442500 127.90 1020.20 G = S/Dt + O/2
1 3850000 149.80 1144.34

Step III : Do the routing calculations using the modified puls relation G2 = G1 + I m − O1 making use of the
Auxiliary curve or by interpolation.

Seifu A. 60
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Time Inflow Out flow

(h) (m3/s) Im Im - O G (m3/s)
0 12 192.28 12
1 42 27 15 207.28 13.22
2 72 57 43.78 251.06 17.67
3 102 87 69.33 320.38 24.83
4 87 94.5 69.67 390.05 33.13
5 72 79.5 46.37 436.42 38.74
6 57 64.5 25.76 462.18 42.15
7 42 49.5 7.35 469.53 43.12
8 27 34.5 -8.62 460.91 41.98
9 12 19.5 -22.48 438.43 39.00
10 12 12 -27.00 411.42 35.68

Step IV : Plot the inflow and outflow hydrographs

Inflow - Outflow Hydrographs

120
100
Discharge (Cu.m/s)

80
Inflow
60
Outflow
40
20
0
0 2 4 6 8 10 12
Tim e (Hr)

Peak outflow discharge = 43.12 m3/s.

Seifu A. 61
Bahir Dar University, ioT School of Civil & Water Resources Engineering

5.3.2 Muskingum Method of Channel Routing

The Muskingum method was developed to directly accommodate the looped relationship between storage
and outflow that exists in rivers. With the Muskingum method, storage within a reach is visualized in two
parts: prism storage and wedge storage. Prism storage is essentially the storage under the steady flow
water surface profile. Wedge storage is the additional storage under the actual water surface profile. As
shown in Figure 5-5 & 5-6, during the rising stages of the floodwave the wedge storage is positive and
added to the prism storage. During the falling stages of a floodwave, the wedge storage is negative and
subtracted from the prism storage.

Figure 5-5 Muskingum prism and wedge storage concepts

Figure 5-6 Looped Storage – Outflow Relationship for a river reach

I) Development of the Muskingum routing equation

Prism storage is computed as the outflow (O) times the travel time through the reach (K). Wedge storage
is computed as the difference between inflow and outflow (I-O) times a weighting coefficient X and the
travel time K. The coefficient K corresponds to the travel time of the floodwave through the reach. The

Seifu A. 62
Bahir Dar University, ioT School of Civil & Water Resources Engineering

parameter X is a dimensionless value expressing a weighting of the relative effects of inflow and outflow
on the storage (S) within the reach. Thus, the Muskingum method defines the storage in the reach as a
linear function of weighted inflow and outflow:
S = prism storage + wedge storage

S = KO + KX ( I − O )

S = K  XI + (1 − X ) O  Eq. 5-5

where
S = total storage in the routing reach O = rate of outflow from the routing reach
I = rate of inflow to the routing reach K = travel time of the floodwave through the reach
X = dimensionless weighting factor, ranging from 0.0 to 0.5

The quantity in the brackets of Equation 5-5 is considered an expression of weighted discharge. When X =
0.0, the equation reduces to S = KO, indicating that storage is only a function of outflow, which is
equivalent to level-pool reservoir routing with storage as a linear function of outflow. When X = 0.5, equal
weight is given to inflow and outflow, and the condition is equivalent to a uniformly progressive wave
that does not attenuate. Thus, “0.0” and “0.5” are limits on the value of X, and within this range the value
of X determines the degree of attenuation of the floodwave as it passes through the routing reach. A value
of “0.0” produces maximum attenuation, and “0.5” produces pure translation with no attenuation.

The Muskingum routing equation is obtained by combining Equation 5-5 with the continuity equation,
Equation 5-1, and solving for O2.

O2 = C1 I 2 + C2 I1 + C3O1 Eq. 5-6

The subscripts 1 and 2 in this equation indicate the beginning and end, respectively, of a time interval t.
The routing coefficients C1, C2, and C3 are defined in terms of ∆t, K, and X.

Eq. 5-8

Eq. 5-9

Eq. 5-10

Given an inflow hydrograph, a selected computation interval ∆t, and estimates for the parameters K and X,
the outflow hydrograph can be calculated.

Seifu A. 63
Bahir Dar University, ioT School of Civil & Water Resources Engineering

II) Determination of Muskingum K and X

a) In a gauged situation, the Muskingum K and X parameters can be calculated from observed inflow and
outflow hydrographs. The travel time, K, can be estimated as the interval between similar points on the
inflow and outflow hydrographs. The travel time of the routing reach can be calculated as the elapsed time
between centroid of areas of the two hydrographs, between the hydrograph peaks, or between midpoints of
the rising limbs. After K has been estimated, a value for X can be obtained through trial and error. Assume
a value for X, and then route the inflow hydrograph with these parameters. Compare the routed
hydrograph with the observed outflow hydrograph. Make adjustments to X to obtain the desired fit.
Adjustments to the original estimate of K may also be necessary to obtain the best overall fit between
computed and observed hydrographs.

b) In an ungauged situation, a value for K can be estimated as the travel time of the floodwave through
the routing reach. The floodwave velocity (Vw) is greater than the average velocity at a given cross section
for a given discharge. The floodwave velocity can be estimated by a number of different techniques:

L
K= Eq. 5-11
Vw

where
Vw = floodwave velocity, in m/second B = top width of the water surface
L = length of the routing reach, in m

One method of estimating floodwave velocity is to estimate the average velocity (V) and multiply it by a
ratio. The average velocity can be calculated from Manning’s equation with a representative discharge and
cross section for the routing reach. For various channel shapes, the floodwave velocity has been found to
be a direct ratio of the average velocity.

Channel shape Ratio Vw/V

Wide rectangular 1.67
Wide parabolic 1.44
Triangular 1.33

For natural channels, an average ratio of 1.5 is suggested.

Once the wave speed has been estimated, the travel time (K) can be calculated with Eq. 5-11.

d) Estimating the Muskingum X parameter in an ungauged situation can be very difficult. X varies
between 0.0 and 0.5, with 0.0 providing the maximum amount of hydrograph attenuation and 0.5 no

Seifu A. 64
Bahir Dar University, ioT School of Civil & Water Resources Engineering

attenuation. Experience has shown that for channels with mild slopes and flows that go out of bank, X will
be closer to 0.0. For steeper streams, with well defined channels that do not have flows going out of bank,
X will be closer to 0.5. Most natural channels lie somewhere in between these two limits, leaving a lot of
room for “engineering judgment.”

One equation that can be used to estimate the Muskingum X coefficient in ungauged areas has been
developed by Cunge (1969). The equation is written as follows:

1 Qo 
X= 1 −  Eq. 5-12
2  BSo c∆x 

where
Qo = reference flow from the inflow hydrograph c = floodwave speed
So = friction slope or bed slope B = top width of the flow area
∆x = length of the routing subreach

The choice of which flow rate to use in this equation is not completely clear. Experience has shown that a
reference flow based on average values (midway between the base flow and the peak flow) is in general
the most suitable choice.

III) Selection of the number of subreaches

The Muskingum equation has a constraint related to the relationship between the parameter K and the
computation interval ∆t. Ideally, the two should be equal, but ∆t should not be less than 2KX to avoid
negative coefficients and instabilities in the routing procedure. Hence,
2KX < ∆t ≤ K Eq. 5-13

A long routing reach should be subdivided into subreaches so that the travel time through each subreach is
approximately equal to the routing interval ∆t. That is:
K
Number of subreaches = Eq. 5-14
∆t
This assumes that factors such as channel geometry and roughness have been taken into consideration in
determining the length of the routing reach and the travel time K.

Seifu A. 65
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Example – Muskingum Routing

The Muskingum constants, K and X, are estimated for a given river reach to be 12 hours and 0.2
respectively. Assuming an initial steady flow (I = O), determine the outflow hydrograph at the
downstream end of the reach for the inflow hydrograph shown in Table below. What is the peak outflow?

Time (hr) 0 6 12 18 24 30 36 42
Inflow (m3/s) 25.0 32.5 58.0 65.0 59.4 49.5 42.5 35.0

Solution:
Muskingum Routing Parameters are given as:
K= 12 hr
X= 0.2
∆t = 6 hr
Check that ∆t < K, i.e., 6 < 12 . OK and ∆t > 2KX, i.e., 6 > 4.8. OK
Step I: Compute the routing coefficients:
C1 = 0.04762
C2 = 0.42857
C3 = 0.52381
SUM 1
Step II: Check that the sum of the coefficients is unity.

Step III : Perform the routing using the equation O2 = C1 I 2 + C2 I1 + C3O1

T (hr) I (m3/s) O (m3/s)

0 25 25.0
6 32.5 25.4
12 58 30.0
18 65 43.7
24 59.4 53.6
30 49.5 55.9
36 42.5 52.5
42 35 47.4

Seifu A. 66
Bahir Dar University, ioT School of Civil & Water Resources Engineering

Step IV : Plot the Inflow & Outflow Hydrographs

The peak discharge at the downstream end is 55.9 m3/s.

Seifu A. 67
Bahir Dar University, ioT School of Civil & Water Resources Engineering

References:
1. Admasu G. (1988). Regional Analysis on some aspects of stream flow characteristics in
Ethiopia (Draft Report)
2. Cunnane, C. (1989). Statistical Distributions for Flood Frequency Analysis. WMO
Operational Hydrology Report No. 33
3. Chow, T.C. (1988). Applied Hydrology. McGraw-Hill series in water resource engineering
and environmental engineering.
4. Dandekar, M.M & Sharma, K.N. (1990). Water Power Engineering.
5. Enku, T. and Melesse, A. (2013) A simple temperature method for the estimation of
evapotranspiration, hydrological processes, DOI: 10.1002/hyp.9844.
6. NMSA. (1996). Climate & agro-climate resources of Ethiopia. NMSA Meteorological
Research Report Series, Vol.1, No.1, Addis Ababa.
7. Peter Smart (2004). Understanding Hydrology, A Comparison of Hydrologic Methods.
8. Ponce, V.M. (1989). Engineering Hydrology: Principles and Practice. Prentice-Hall plc.
9. Shaw, E.M. (1994). Hydrology in Practice
10. Walker ,S.E et al . Application of the SCS Curve Number Method to Mildly-Sloped
Watersheds
11. Varshney, R.S.(1979). Engineering Hydrology, 2nd Ed.
12. USACE (1993). Hydrologic Frequency Analysis, Engineering Manual, EM 1110-2-1415.
13. USACE (1994). Flood Runoff Analysis, Engineering Manual, EM 1110-2-1417 .

Seifu A. 68