Advanced hydrology
-
Upload
jisha-john -
Category
Engineering
-
view
3.755 -
download
18
Transcript of Advanced hydrology
Advanced Hydrology(Web course)
Subhankar KarmakarAssistant Professor
Centre for Environmental Science and Engineering (CESE)Indian Institute of Technology Bombay
Powai, Mumbai 400 076Email: [email protected]
Ph. # +91 22 2576 7857
Hydrologic Cycle
Prof. Subhankar KarmakarIIT Bombay
Module 13 Lectures
The objective of this module is to introduce the
phenomena of weather, different stages of the hydrologic
cycle, hydrologic losses and its measurements.
Module 1
Topics to be covered
Weather
Introduction to Hydrology
Different stages of Hydrology or water cycle
Hydrologic losses and measurements
Analytical Methods
Empirical Methods
Module 1
Lecture 1: Weather and hydrologic cycle
Module 1
Weather & Climate
Weather- “the state of the atmosphere with respect to heat or cold, wetness
or dryness, calm or storm, clearness or cloudiness”.
Climate – “the average course or condition of the weather at a place usually
over a period of years as exhibited by temperature, wind velocity, and
precipitation”.
(Wikipedia)
Weather refers, generally, to day-to-day temperature and precipitation activity, whereas climate is the term for the average atmospheric conditions over longer periods of time.
Module 1Lecture 1
Atmosphere
Troposphere Most of the weather occurs.
Stratosphere19% of the atmosphere’s gases; Ozone layer
Mesosphere Most meteorites burn up here.
Thermosphere High energy rays from the sun are absorbed; Hottest layer.
Exosphere Molecules from atmosphere escape into space; satellites orbit here.
(http://www.windows.ucar.edu/tour/link=/earth/Atmosphere/layers_activity_print.html) Module 1Lecture 1
Winds and Wind belts
Exist to circulate heat and
moisture from areas of heating
to areas of cooling
Equator to poles
Low altitudes to high
altitudes
Three bands of low and high
pressure above and below the
equator (area of low pressure)
Module 1Lecture 1
Cloud Types
Cloud is a visible set of drops of water and fragments of ice suspended inthe atmosphere and located at some altitude above the earth’s surface.
Module 1Lecture 1
Classification of Precipitation events
Based on the “mechanism” by which air is lifted.
Frontal lifting:
Warmer air is forced to go above cooler air in equilibrium with a cooler surface.
Orographic lifting:
Air is forced to go over mountains (and it’s the reason why windward slopes
receive more precipitation).
Convective Lifting:
Warm air rises from a warm surface and progressively cools down.
Cyclonic Lifting:
A cyclonic storm is a large, low pressure system that forms when a warm air
mass and a cold air mass collide.
Module 1Lecture 1
Frontal lifting
Module 1Lecture 1
Orographic lifting
Module 1Lecture 1
Convectional lifting
(climateofindia.pbworks.com)
Module 1Lecture 1
Cyclonic lifting
Module 1Lecture 1
Factors affecting Indian climate
Related to Location and Relief Related to Air Pressure and Wind
•Latitude
•Altitude
•Relief
•Distance from Sea
•The Himalayan Mountains
•Distribution of Land & water
•Surface pressure & wind
•Upper air circulation
•Western cyclones
Module 1
Factors affecting Indian climate
Lecture 1
Module 1
Seasons
Cold weather
Hot weather
South west monsoon
Retreating monsoon
Lecture 1
► It extends from December toFebruary.
► Vertical sun rays shift towardssouthern hemisphere.
► North India experiencesintense cold
► Light wind blow makes thisseason pleasant in southIndia.
► Occasional tropical cyclonevisit eastern coast in thisseason.
Tropical Cyclone
Cold Weather Season
Seasons
Module 1Lecture 1
250C
250C
200C
200C200C
150C
200C
100C`
Temperature-January
(climateofindia.pbworks.com) Module 1
Seasons
Lecture 1
Pressure-January
(climateofindia.pbworks.com)
1014
HIGH PRESSURE
Module 1
Seasons
Lecture 1
RAINFALL DUE TO WESTERN
DISTURBANCES
RAINFALL DUE TO NORTH EAST
WIND
Winter Rainfall
Module 1
Seasons
Lecture 1
► It extends from March to May.
► Vertical sun rays shift towards Northern hemisphere.
► Temperature rises gradually from south to north.
► Highest Temperature experiences in Karnataka in March, Madhya Pradesh in April and Rajastan in May. March 300C
April 380C
May 480C
Hot Weather Season
Module 1
Seasons
Lecture 1
Temperature-July
250C
300C
Module 1
Seasons
Lecture 1
Pressure-July
Module 1
Seasons
Lecture 1
LOO
KALBAISAKHI
BARDOLI CHHEERHA
MANGO SHOWER
BLOSSOM SHOWER
Storms in Hot Weather Season
(climateofindia.pbworks.com) Module 1
Seasons
Lecture 1
► It extends from June to September.
► Intense heating in north west India creates low pressure region.
► Low pressure attract the wind from the surrounding region.
► After having rains for a few days sometime monsoon fails to occur for one or more weeks is known as break in the monsoon.
South West Monsoon
LOW PRESSURE HIGH TEMPERATURE
Module 1
Seasons
Lecture 1
INTER TROPICAL CONVERGENCE ZONE
Arabian sea Branch
Bay of Bengal Branch
Monsoon Wind
Module 1
Seasons
Lecture 1
Onset of SW Monsoon
Module 1
Seasons
Lecture 1
► It extends from October to November
► Vertical sun rays start shifting towards Northern hemisphere.
► Low pressure region shift from northern parts of India towards south.
► Owing to the conditions of high temperature and humidity, the weather becomes rather oppressive. This is commonly known as the ‘October heat’
LOW PRESSURE
Retreating Monsoon Season
Module 1
Seasons
Lecture 1
Withdrawal of Monsoon
Module 1
Seasons
Lecture 1
> 200cm
100-200cm
50-100 cm
< 50cm
Distribution of Rainfall
(climateofindia.pbworks.com) Module 1
Seasons
Lecture 1
► The variability of rainfall is computed with the help of the following formula: C.V.= Standard Deviation/ Mean * 100
► Variability <25% exist in Western coasts, Western Ghats, north-eastern peninsula, eastern plain of the Ganga, northern-India, Uttaranchal, SW J & K & HP.
► Variability >50% found in Western Rajastan, J & K and interior parts of Deccan.
► Region with high rainfall has less variability.
Variability of Rainfall
Module 1
Seasons
Lecture 1
Lecture 2: Weather and hydrologic cycle (contd.)
Module 1
Hydrology
Hydor + logos (Both are Greek words)
“Hydor” means water and “logos” means study.
Hydrology is a science which deals with the occurrence, circulation and
distribution of water of the earth and earth’s atmosphere.
Hydrological Cycle: It is also known as water cycle. The hydrologic cycle is a
continuous process in which water is evaporated from water surfaces and the
oceans, moves inland as moist air masses, and produces precipitation, if the
correct vertical lifting conditions exist.
Module 1Lecture 1
Hydrologic Cycle
(climateofindia.pbworks.com) Module 1Lecture 1
Stages of the Hydrologic cycle
Precipitation
Infiltration
Interception
Depression storage
Run-off
Evaporation
Transpiration
Groundwater
Module 1Lecture 1
Forms of precipitation
Rain
Water drops that have a diameter of at least 0.5 mm. It can be classified based on
intensity as,
Light rain up to 2.5 mm/h
Moderate rain2.5 mm/h to 7.5 mm/h
Heavy rain > 7.5 mm/h
Snow
Precipitation in the form of ice crystals which usually combine to form flakes, with
an average density of 0.1 g/cm3.
Drizzle
Rain-droplets of size less than 0.5 mm and rain intensity of less than 1mm/h is
known as drizzle.
Precipitation
Module 1Lecture 1
Glaze
When rain or drizzle touches ground at 0oC, glaze or freezing rain is
formed.
Sleet
It is frozen raindrops of transparent grains which form when rain falls
through air at subfreezing temperature.
Hail
It is a showery precipitation in the form of irregular pellets or lumps of ice of
size more than 8 mm.
Precipitation
Module 1
Forms of precipitation Contd…
Lecture 1
Rainfall measurement
The instrument used to collect and measure the precipitation is called raingauge.
Types of raingauges:
1) Non-recording : Symon’s gauge
2) Recording
Tipping-bucket type
Weighing-bucket type
Natural-syphon type
Symon’s gauge
Precipitation
Module 1Lecture 1
Recording raingauges
The instrument records the graphical variation of the rainfall, the total
collected quantity in a certain time interval and the intensity of the rainfall
(mm/hour).
It allows continuous measurement of the rainfall.
Precipitation
1. Tipping-bucket type
These buckets are so balanced that when
0.25mm of rain falls into one bucket, it tips
bringing the other bucket in position.
Tipping-bucket type raingauge
Module 1Lecture 1
Precipitation
2. Weighing-bucket type
The catch empties into a bucket mounted
on a weighing scale.
The weight of the bucket and its contents
are recorded on a clock work driven chart.
The instrument gives a plot of cumulative
rainfall against time (mass curve of
rainfall).
Weighing-bucket type raingauge
Module 1Lecture 1
Precipitation
3. Natural Syphon Type (Float Type)
The rainfall collected in the funnel
shaped collector is led into a float
chamber, causing the float to rise.
As the float rises, a pen attached to
the float through a lever system
records the rainfall on a rotating drum
driven by a clockwork mechanism.
A syphon arrangement empties the
float chamber when the float has
reached a preset maximum level.
Float-type raingauge
Module 1Lecture 1
Hyetograph
Plot of rainfall intensity against
time, where rainfall intensity is depth of
rainfall per unit time
Mass curve of rainfall
Plot of accumulated precipitation
against time, plotted in chronological
order.
Point rainfall
It is also known as station rainfall . It
refers to the rainfall data of a station
Presentation of rainfall data
Rainfall Mass Curve
Precipitation
Module 1Lecture 1
The following methods are used to measure the average precipitation
over an area:
1. Arithmetic Mean Method
2. Thiessen polygon method
3. Isohyetal method
4. Inverse distance weighting
Precipitation
Mean precipitation over an area
1. Arithmetic Mean MethodSimplest method for determining areal average
where, Pi : rainfall at the ith raingauge station N : total no: of raingauge stations
∑=
=N
iiP
NP
1
1
P1
P2
P3
Module 1Lecture 1
2. Thiessen polygon method
This method assumes that any point in the watershed receives the same
amount of rainfall as that measured at the nearest raingauge station.
Here, rainfall recorded at a gage can be applied to any point at a distance
halfway to the next station in any direction.
Steps:
a) Draw lines joining adjacent gages
b) Draw perpendicular bisectors to the lines created in step a)
c) Extend the lines created in step b) in both directions to form representative
areas for gages
d) Compute representative area for each gage
Module 1
Precipitation
Mean precipitation over an area Contd…
Lecture 1
e) Compute the areal average using the following:
∑=
=N
iii PA
AP
1
1
mmP 7.2047
302020151012=
×+×+×=
P1
P2
P3
A1
A2
A3
P1 = 10 mm, A1 = 12 Km2
P2 = 20 mm, A2 = 15 Km2
P3 = 30 mm, A3 = 20 km2
3. Isohyetal method
∑=
=N
iii PA
AP
1
1
mmP .2150
35102515152055 =×+×+×+×
=
where, Ai : Area between each pair of adjacent isohyets
Pi : Average precipitation for each pair of
adjacent isohyets
P2
10
20
30
A2=20 , p2 = 15
A4=10 , p3 = 35
P1
P3
A1=5 , p1 = 5
A3=15 , p3 = 25
Steps:
a) Compute distance (di) from ungauged point
to all measurement points.
b) Compute the precipitation at the ungauged
point using the following formula:
N = No: of gauged points
4. Inverse distance weighting (IDW) method
Prediction at a point is more influenced by nearby measurements than that
by distant measurements. The prediction at an ungauged point is inversely
proportional to the distance to the measurement points.
( ) ( )2212
2112 yyxxd −+−=
P1=10
P2= 20
P3=30
d1=25
d2=15
d3=10p
∑
∑
=
=
=N
i i
N
i i
i
d
dP
P
12
12
1ˆ mmP 24.25
101
151
251
1030
1520
2510
ˆ
222
222=
++
++=
Module 1Lecture 1
Check for continuity and consistency of rainfall records
Normal rainfall as standard of comparison
Normal rainfall: Average value of rainfall at a particular date, month or yearover a specified 30-year period.
Adjustments of precipitation data
Check for Continuity: (Estimation of missing data)
P1, P2, P3,…, Pm annual precipitation at neighboring M stations 1, 2, 3,…, M
respectively
Px Missing annual precipitation at station X
N1, N2, N3,…, Nm & Nx normal annual precipitation at all M stations and at X
respectively
Precipitation
Module 1Lecture 1
Check for continuity
1. Arithmetic Average Method:This method is used when normal annual precipitations at various
stations show variation within 10% w.r.t station X
2. Normal Ratio Method
Used when normal annual precipitations at various stations show
variation >10% w.r.t station X
Precipitation
Module 1
Adjustments of precipitation data Contd…
Lecture 1
Test for consistency of record
Causes of inconsistency in records:
Shifting of raingauge to a new location
Change in the ecosystem due to calamities
Occurrence of observational error from a certain date
Relevant when change in trend is >5years
Precipitation
Module 1
Adjustments of precipitation data Contd…
Lecture 1
Double Mass Curve Technique
Acc
umul
ated
Pre
cipi
tatio
n of
St
atio
n X,
ΣPx
Average accumulated precipitation of neighbouring stations ΣPav
9089
88
87
86
85
84
8382
When each recorded data comes
from the same parent population, they
are consistent.
Break in the year : 1987 Correction Ratio : Mc/Ma = c/a Pcx = Px*Mc/Ma
Pcx – corrected precipitation at any time period t1 at station XPx – Original recorded precipitation at time period t1 at station XMc – corrected slope of the double mass curveMa – original slope of the mass curve
Module 1
Precipitation
Adjustments of precipitation data Contd…
Test for consistency of record
Lecture 1
It indicates the areal distribution characteristic of a storm of given duration.
Depth-Area relationship
For a rainfall of given duration, the average depth decreases with the area in
an exponential fashion given by:
where : average depth in cms over an area A km2,
Po : highest amount of rainfall in cm at the storm centre
K, n : constants for a given region
Precipitation
Depth-Area-Duration relationships
)exp(0nKAPP −=
P
Module 1Lecture 1
The development of maximum depth-area-duration relationship is known
as DAD analysis.
It is an important aspect of hydro-meteorological study.
Typical DAD curves(Subramanya, 1994)
Module 1
Precipitation
Depth-Area-Duration relationships Contd…
Lecture 1
It is necessary to know the rainfall intensities of different durations and different
return periods, in case of many design problems such as runoff
disposal, erosion control, highway construction, culvert design etc.
The curve that shows the inter-dependency between i (cm/hr), D (hour) and T
(year) is called IDF curve.
The relation can be expressed in general form as:
( )n
x
aDTki+
=i – Intensity (cm/hr)
D – Duration (hours)
K, x, a, n – are constant for a given catchment
Intensity-Duration-Frequency (IDF) curves
Precipitation
Module 1Lecture 1
0
2
4
6
8
10
12
14
0 1 2 3 4 5 6
Inte
nsi
ty, c
m/h
r
Duration, hr
Typical IDF Curve
T = 25 years
T = 50 years
T = 100 years
k = 6.93x = 0.189a = 0.5n = 0.878
Module 1
Precipitation
Intensity-Duration-Frequency (IDF) curves Contd…
Lecture 1
Exercise Problem
• The annual normal rainfall at stations A,B,C and D in a basin are 80.97,
67.59, 76.28 and 92.01cm respectively. In the year 1975, the station D was
inoperative and the stations A,B and C recorded annual precipitations of
91.11, 72.23 and 79.89cm respectively. Estimate the rainfall at station D in
that year.
Precipitation
Module 1Lecture 1
Lecture 3: Hydrologic losses
Module 1
In engineering hydrology, runoff is the main area of interest. So, evaporation
and transpiration phases are treated as “losses”.
If precipitation not available for surface runoff is considered as “loss”, then the
following processes are also “losses”:
Interception
Depression storage
Infiltration
In terms of groundwater, infiltration process is a “gain”.
Hydrologic losses
Module 1Lecture 2
Interception is the part of the rainfall that is intercepted by the earth’s surface
and which subsequently evaporates.
The interception can take place by vegetal cover or depression storage in
puddles and in land formations such as rills and furrows.
Interception can amount up to 15-50% of precipitation, which is a significant part
of the water balance.
Interception
Module 1Lecture 2
Depression storage is the natural depressions within a catchment area which
store runoff. Generally, after the depression storage is filled, runoff starts.
A paved surface will not detain as much water as a recently furrowed field.
The relative importance of depression storage in determining the runoff from a
given storm depends on the amount and intensity of precipitation in the storm.
Depression storage
Module 1Lecture 2
Infiltration
The process by which water on the ground surface enters the soil. The rate of
infiltration is affected by soil characteristics including ease of entry, storage
capacity, and transmission rate through the soil.
The soil texture and structure, vegetation types and cover, water content of the
soli, soil temperature, and rainfall intensity all play a role in controlling
infiltration rate and capacity.
Module 1Lecture 2
Infiltration capacity or amount of infiltration
depends on :
Soil type
Surface of entry
Fluid characteristics.
http://techalive.mtu.edu/meec/module01/images/Infiltration.jpg
Infiltration
Factors affecting infiltration
Module 1Lecture 2
Soil Type : Sand with high porosity will have greater infiltration than clay soil with
low porosity.
Surface of Entry : If soil pores are already filled with water, capacity of the soil to
infiltrate will greatly reduce. Also, if the surface is covered by leaves or impervious
materials like plastic, cement then seepage of water will be blocked.
Fluid Characteristics : Water with high turbidity or suspended solids will face
resistance during infiltration as the pores of the soil may be blocked by the
dissolved solids. Increase in temperature can influence viscosity of water which will
again impact on the movement of water through the surface.
Infiltration
Module 1
Factors affecting infiltration Contd…
Lecture 2
Infiltration
Infiltration capacity :
The maximum rate at which, soil at a given time can absorb water.
f = fc when i ≥ fc
f = when i < fcwhere fc = infiltration capacity (cm/hr)
i = intensity of rainfall (cm/hr)
f = rate of infiltration (cm/hr)
Module 1
Infiltration rate
Lecture 2
Infiltration
Horton’s Formula: This equation assumes an infinite water supply at the surface
i.e., it assumes saturation conditions at the soil surface.
For measuring the infiltration capacity the following expression are used:
f(t) = fc + (f0 – fc) e–kt for where k = decay constant ~ T-1
fc = final equilibrium infiltration capacity
f0 = initial infiltration capacity when t = 0
f(t) = infiltration capacity at any time t from start of the rainfall
td = duration of rainfall
Module 1
Infiltration rate Contd…
Lecture 2
f0
ft=fc+(f0-fc)e -kt
fcf
infiltration
time t
Infiltration
Graphical representation of Horton formula
Measurement of infiltration1. Flooding type infiltrometer
2. Rainfall simulator
Module 1
Infiltration rate Contd…
Lecture 2
Infiltration
Infiltration indices
The average value of infiltration is called
infiltration index.
Two types of infiltration indices
φ - index
w –index
Module 1
Measurement of infiltration
Lecture 2
Infiltration
The indices are mathematically expressed as:
where P=total storm precipitation (cm)
R=total surface runoff (cm)
Ia=Initial losses (cm)
te= elapsed time period (in hours)
The w-index is more accurate than the φ-index because it subtracts initial losses
φ-index=(P-R)/te
w-index=(P-R-Ia)/te
Module 1
Measurement of infiltration Contd...
Infiltration indices
Lecture 2
Example Problem
Infiltration
A 12-hour storm rainfall with the following depths in cm occurred over a basin:
2.0, 2.5, 7.6, 3.8, 10.6, 5.0, 7.0, 10.0, 6.4, 3.8, 1.4 and 1.4. The surface runoff
resulting from the above storm is equivalent to 25.5 cm of depth over the basin.
Determine the average infiltration index (Φ-index) for the basin.
Total rainfall in 12 hours = 61.5 cm
Total runoff in 12 hours = 25.5 cm
Total infiltration in 12 hours = 36 cm
Average infiltration = 3.0 cm/hr
Average rate of infiltration during the central 8 hours
8 Φ +2.0+2.5+1.4+1.4 = 36
Φ = 3.6cm/hr
Module 1Lecture 2
In this process, water changes from its liquid state to gaseous state.
Water is transferred from the surface to the atmosphere
through evaporation
Evaporation
Evaporation is directly proportional to :
Vapor pressure (ew),
Atmospheric temperature (T),
Wind speed (W) and
Heat storage in the water body (A)
Module 1Lecture 2
Evaporation
Vapour pressure: The rate of evaporation is proportional to the difference
between the saturation vapour pressure at the water temperature, ew and the
actual vapour pressure in the air ea.
EL = C (ew - ea)EL = rate of evaporation (mm/day); C = a constant ; ew and ea are in mm of
mercury;
The above equation is known as Dalton’s law of evaporation. Evaporation takes
place till ew > ea, condensation happen if ew < ea
Module 1
Factors affecting evaporation
Lecture 2
Temperature: The rate of evaporation increase if the water temperature is
increased. The rate of evaporation also increase with the air temperature.
Heat Storage in water body: Deep bodies can store more heat energy than
shallow water bodies. Which causes more evaporation in winter than summer
for deep lakes.
Evaporation
Module 1
Factors affecting evaporation Contd…
Lecture 2
Soil evaporation: Evaporation from water stored in the pores of the soil i.e., soil
moisture.
Canopy evaporation: Evaporation from tree canopy.
Total evaporation from a catchment or an area is the summation of both soil
and canopy evaporation.
Evaporation
Module 1
Types of Evaporation
Lecture 2
Evaporation
The amount of water evaporated from a water surface is estimated by the
following methods:
1. Using evaporimeter data
2. Empirical equations
3. Analytical methods
1. Evaporimeters : Water containing pans which are exposed to the atmosphere
and loss of water by evaporation measured in them in the regular intervals.
a) Class A Evaporation Pan
b) ISI Standard pan
c) Colorado sunken pan
d) USGS Floating pan
Measurement of evaporation
Module 1Lecture 2
Evaporation
Demerits of Evaporation pan:
1. Pan differs in the heat-storing capacity and heat transfer from the sides
and bottom.
Result: reduces the efficiency (sunken pan and floating pan eliminates this
problem)
2. The height of the rim in an evaporation pan affects the wind action over the
surface.
3. The heat-transfer characteristics of the pan material is different from that of
the reservoir.
Module 1
Measurement of evaporation Contd…
1. Evaporimeters
Lecture 2
Evaporation
Pan Coefficient (Cp)For accurate measurements from evaporation pan a coefficient is introduce, known
as pan coefficient (Cp). Lake evaporation = Cp x pan evaporation
Type of pan Range of Cp Average value Cp
Class A land pan 0.60-0.80 0.70
ISI pan 0.65-1.10 0.80
Colorado sunken pan 0.75-0.86 0.78
USGS Floating pan 0.70-0.82 0.80
Source: Subramanya, 1994
Module 1
Measurement of evaporation Contd…
Lecture 2
2. Empirical equation
Mayer’s Formula (1915)
EL = Km (ew- ea) (1+ (u9/16))
where EL = Lake evaporation in mm/day;
ew = saturated vapour pressure at the water surface temperature;
ea = actual vapour pressure of over lying air at a specified height;
u9 = monthly mean wind velocity in km/hr at about 9 m above the
ground;
Km= coefficient, 0.36 for large deep waters and 0.50 for small
shallow waters.
Evaporation
Module 1
Measurement of evaporation Contd…
Lecture 2
A reservoir with a surface area of 250 ha had the following parameters: water temp.
22.5oC, RH = 40%, wind velocity at 9.0 m above the ground = 20 km/hr. Estimate
the volume of the water evaporated from the lake in a week.
Given ew = 20.44, Km =0.36.
Solution:
ea = 0.40 x 20.44 = 8.176 mm Hg; U9 = 20 km/hr;
Substitute the values in Mayer’s Equation .
Now, EL = 9.93 mm/day
For a week it will be 173775 m3.
Evaporation
Example Problem
Module 1Lecture 2
Water Budget method: This is the simplest analytical method.
P + Vis + Vig = Vos + EL + ds + TL
P= daily precipitation;
Vis = daily surface inflow into the lake;
Vig = daily groundwater flow ;
Vos= daily surface outflow from the lake;
Vog= daily seepage outflow;
EL= daily lake evaporation;
ds= increase the lake storage in a day;
TL= daily transportation loss
3. Analytical method
Module 1
Evaporation
Measurement of evaporation Contd…
Lecture 2
Evapotranspiration
Transpiration + Evaporation
This phenomenon describes transport of water into the atmosphere from
surfaces, including soil (soil evaporation), and vegetation (transpiration).
Hydrologic Budget equation for Evapotranspiration:
P – Rs – Go - Eact = del S
P= precipitation; Rs= Surface runoff; Go= Subsurface outflow; Eact = Actual
evapotranspiration; del S = change in the moisture storage.
Module 1Lecture 2
Highlights in the Module
Hydrology is a science which deals with the movement, distribution, and qualityof water on Earth including the hydrologic cycle, water resources andenvironmental watershed sustainability.
Stages of the Hydrologic cycle or Water cycle Precipitation Infiltration Interception Run-off Evaporation Transpiration Groundwater
Module 1
Hydrologic Losses : evaporation, transpiration and interception
Measurement of Precipitation
Non-Recording Rain gauges: Symons’s gauge
Recording Rain gauges: tipping bucket type, weighing bucket type and natural
syphon type
Presentation of Rainfall Data: Mass curve, Hyetograph, Point Rainfall and DAD
curves
Factors affecting Infiltration: soil characteristics, surface of entry and fluid characteristics
Determination of Infiltration rate can be performed using flooding type infiltrometers and rainfall simulator.
Module 1
Highlights in the Module Contd…
Factors affecting evaporation : vapour pressure, wind speed, temperature, atmospheric pressure, presence of soluble salts and heat storage capacity of lake/reservoir
Measurement of evaporation: evaporimeters, empirical equations and analytical methods
Weather refers, generally, to day-to-day temperature and precipitationactivity, whereas climate is the term for the average atmospheric conditionsover longer periods of time.
Formation of Precipitation: frontal, convective, cyclonic and orographic
The four different seasons are: Cold weather, Hot weather, South-Westmonsoon and Retreating monsoon
Module 1
Highlights in the Module Contd…
Prof. Subhankar KarmakarIIT Bombay
Philosophy of Mathematical Models of Watershed Hydrology
Module 22 Lectures
Philosophy of mathematical models of watershed hydrology
Lecture 1
Objectives of this module is to introduce the terms andconcepts in mathematical modelling which will form as a toolfor effective and efficient watershed management throughwatershed modelling
Module 2
Topics to be covered
Concept of mathematical modeling
Watershed - Systems Concept
Classification of Mathematical Models
Different Components in Mathematical Modelling
Module 2
A model is a representation of reality in simple form based on hypotheses and equations:
There are two types of models Conceptual Mathematical
Modeling Philosophy
Experiment
Computation
Theory
Module 2
Conceptual Models
Qualitative, usually based on graphs
Represent important system:
components
processes
linkages
Interactions
Conceptual Models can be used:
As an initial step
For hypothesis testing
For mathematical model development
As a framework
For future monitoring, research, and management actions at a site
Modeling = The use of mathematics as a tool to explain and make predictions of natural phenomena (Cliff Taubes, 2001)
Mathematical modelling may involve words, diagrams, mathematical notation and physical structure
This aims to gain an understanding of science through the use of mathematical models on high performance computers
Science
MathematicsComputer Science
Module 2
Mathematical Models
Mathematical modeling of watershed can address a wide range of environmental
and water resources problems.
Planning, designing and managing water resources systems involve impact
prediction which requires modelling.
Developing a model is an art which requires knowledgeof the system being modeled, the user’s
objectives, goals and information needs, and some analytical and programming skills.
(UNESCO, 2005)
Module 2
Mathematical Models Contd…
Mathematical Modeling Process
Working Model
Mathematical Model
Computational Model
Results/
Conclusions
Real World Problem
Simplify Represent
Translate
Simulate
Interpret
Module 2
Mean – average or expected value
Variance – average of squared deviations from the mean value
Reliability – Probability (satisfactory state)
Resilience – Probability (satisfactory state following unsatisfactory state)
Robustness – adaptability to other than design input conditions
Vulnerability – expected magnitude or extent of failure when
unsatisfactory state occurs
Consistency- Reliability or uniformity of successive results or events
Module 2
Overall measures of system performance
Watershed - Systems Concept
Input Output(Eg. Rainfall, Snow etc.)
(Eg. Discharge)
http://www.desalresponsegroup.org/alt_watershedmgmt.html
Module 2
The Modeling Process
Model World
Mathematical Model(Equations)
Real World
Input parameters
Interpret and Test(Validate) Formulate
Model World Problem
Model Results
Mathematical Analysis
Solutions,Numericals
Module 2
Model:
A mathematical description of the watershed system.
Model Components:
Variables, parameters, functions, inputs, outputs of the watershed.
Model Solution Algorithm:
A mathematical / computational procedure for performing operations on the model for getting outputs from inputs of a watershed.
Types of Models Descriptive (Simulation)
Prescriptive (Optimization)
Deterministic
Probabilistic or Stochastic
Static
Dynamic
Discrete
Continuous
Deductive, inductive, or floating
Basic Concepts
Module 2
Categories of Mathematical Models
TypeEmpirical
Based on data analysisMechanistic
Mathematical descriptions based on theory
Time FactorStatic or steady-state
Time-independentDynamic
Describe or predict system behavior over time
Treatment of Data Uncertainty and VariabilityDeterministic
Do not address data variabilityStochastic
Address variability/uncertainty
Module 2
Classification of Watershed Models
Based on nature of the algorithms
Empirical
Conceptual
Physically based
Based on nature of input and uncertainty
Deterministic
Stochastic
Based on nature of spatial representation
Lumped
Distributed
Black-box
Module 2
Based on type of storm event
Single event
Continuous event
It can also be classified as:
Physical models
Hydrologic models of watersheds;
Scaled models of ships
Conceptual
Differential equations,
Optimization
Simulation modelsModule 2
Classification of Watershed Models Contd…
Descriptive:
That depicts or describes how things actually work, and answers the
question, "What is this?“
Prescriptive:
suggest what ought to be done (how things should work) according to an
assumption or standard.
Deterministic:
Here, every set of variable states is uniquely determined by parameters in the
model and by sets of previous states of these variables. Therefore, deterministic
models perform the same way for a given set of initial conditions.
Module 2
Classification of Watershed Models Contd…
Probabilistic (stochastic):In a stochastic model, randomness is present, and variable states are not describedby unique values, but rather by probability distributions.
Static:A static model does not account for the element of time, while a dynamic modeldoes.
Dynamic:Dynamic models typically are represented with difference equations or differentialequations.
Discrete:A discrete model does not take into account the function of time and usually usestime-advance methods, while a Continuous model does.
Module 2
Classification of Watershed Models Contd…
Deductive, inductive, or floating: A deductive model is a logical structure based on
a theory. An inductive model arises from empirical findings and generalization from
them. The floating model rests on neither theory nor observation, but is merely the
invocation of expected structure.
Single event model:
Single event model are designed to simulate individual storm events and have no
capabilities for replenishing soil infiltration capacity and other watershed abstraction.
Continuous:
Continuous models typically are represented with f(t) and the changes are reflected
over continuous time intervals.
Module 2
Classification of Watershed Models Contd…
Black Box Models:
These models describe mathematically the relation between rainfall and surface
runoff without describing the physical process by which they are related. e.g. Unit
Hydrograph approach
Lumped models:
These models occupy an intermediate position between the distributed models and
Black Box Models. e.g. Stanford Watershed Model
Distributed Models:
These models are based on complex physical theory, i.e. based on the solution of
unsteady flow equations.Module 2
Classification of Watershed Models Contd…
Watershed Modelling Terminology
Input variablesspace-time fields of precipitation, temperature, etc.
Parameters Size Shape Physiography Climate Hydrogeology Socioeconomics
State variablesspace-time fields of soil moisture, etc.
Drainage Land use Vegetation Geology and Soils Hydrology
Module 2
Equations variables
Independent variablesspace x
time t
Dependent variablesdischarge Q
water level h
All other variables are function of the independent or dependent
variables
Module 2
Watershed Modelling Terminology Contd…
Goals & Objectives
Both goals and objectives are very important to accomplish a project. Goals without
objectives can never be accomplished while objectives without goals will never take
you to where you want to be.
Goals Objectives
Vague, less structured Very concrete, specific and measurable
High level statements that provide
overall context of what the project is
trying to accomplish
Attainable, realistic and low level
statements that describe what the project
will deliver.
Tangible
Intangible
Long term
Short termGoals
Module 2
Philosophy of mathematical models of watershed hydrology (contd.)
Lecture 2
time
Prec
ipita
tion
time
flow
Hydrologic Model
The goal considered here is to simulate the shape of a hydrograph given a known input (Eg: rainfall)
Watershed
I. Goal
Module 2
Watershed Modeling Methodology
II. Conceptualization
Source: Wurbs and James, 2002
The hydrologic cycle is a conceptual model that describes the storage andmovement of water between the biosphere, atmosphere, lithosphere, and thehydrosphere.
Module 2
Note:For 90 yrs of record,(2/3) of 90 = 60 yrs for calibrationRemaining (1/3) of 90 = 30 yrs for validation
III. Model Formulation
Hypothetical data
Goals and Objectives
Conceptualization
Model Formulation
Conceptual Representation
Calibration & Verification
Validation
Good
Sensitivity Analysis
Yes
No
Final ModelModule 2
IV. Conceptual Representation
Un measured
Disturbances
Measured
Disturbances
Process State Variable
Eg: velocity, discharge etc.
( )txc ,
Measured errors
System Response
Processed Output
( )txc ,0
•Hypothetical data is considered for sensitivity analysis•Field data is not necessary
Precipitation
Interception Storage
Surface Runoff
Groundwater Storage
Channel Processes
InterflowDirect Runoff
Surface Storage
BaseflowPercolation
Infiltration
ET
ET
(McCuen, 1989)
ET: evapo-transpiration
Module 2
Flowchart of simple watershed model
IV. Conceptual Representation Contd…
Target Modelling
Data Availability
Complexity of Representation
Guidelines for the Conceptual Model
Eg: Flood event
•Spatial data,
•Time series vs events,
•Surrogate data,
•Heterogeneity of basin characteristics
Issues:
•Catchment scale,
•Accuracy of the analysis,
•Computational aspects
To develop a conceptual watershed model, the following inter-related components are to be dealt with:
IV. Conceptual Representation Contd…
Calibration is the activity of verifying that a model of a given problem in a specified
domain correctly describes the phenomena that takes place in that domain.
During model calibration, values of various relevant coefficients are adjusted in
order to minimize the differences between model predictions and actual observed
measurements in the field.
Verification is performed to ensure that the model does what it is intended to do.
V. Calibration & Verification
Module 2
Validation is performed using some other dataset (that has not been used as
dataset for calibration)
It is the task of demonstrating that the model is a reasonable representation of the
actual system so that it reproduces system behaviour with enough fidelity to satisfy
analysis objectives.
For most models there are three separate aspects which should be considered
during model validation:
Assumptions
Input parameter values and distributions
Output values and conclusions
VI. Validation
Module 2
VII. Sensitivity Analysis
Change inputs or parameters, look at the model results
Sensitivity analysis checks the relationships
Sensitivity Analysis
Automatic
Trial & Error•Change input data and re-solve the problem•Better and better solutions can be discovered
Module 2
Sensitivity is the rate of change in one factor with respect to change in another
factor.
A modeling tool that can provide a better understanding of the relation between
the model and the physical processes being modeled.
Let the parameters be and system output be ( )txc ,0β
I Modelβ1β2β3β4
C1C2C3
j
j
i
i
ijC
C
Sβ
β∆
∆
=Sensitivity of ith output to change in jth parameter:i = 1, 2, 3; j = 1,2,3,4
VII. Sensitivity Analysis Contd…
Module 2
( ) ( )
( ) ( ) jijjijii
n
j jii
n
i ijj
CCC
nCCnHere
βββ
ββ
−=∆−=∆
== ∑∑ ==
;
; 11
Mod
el O
utpu
t
Observed Value
x
xx
x
xx
x
xx
x
x
x
x
x
x In this range, model is not good
A straight line indicates an ‘excellent’ model
‘A reasonably good model’
Module 2
VII. Sensitivity Analysis Contd…
Mod
el O
utpu
t
Observed Value
x
xx
x
xx
x
xx
x
x
x
x
x
x
The model may become crude if the system suddenly changes and the model does not incorporate the relevant changes occurred.
A ‘Crude Model’
xx
x
xx
x
xx
x xx
x
xx
x
Module 2
VII. Sensitivity Analysis Contd…
Highlights in the Module
Mathematical modelling may involve words, diagrams, mathematical notation and
physical structure
Mathematical modeling of watershed can address a wide range of environmental
and water resources problems
Different Components in Modelling are:
1)Goals and Objectives,
2) Conceptualization,
3) Model formulation,
4) Sensitivity Analysis,
5) Conceptual Representation,
6) Calibration & Verification,
7) Validation
Module 2
There are different measures of system performance of models:
Mean,
Variance,
Reliability,
Resilience,
Robustness,
Vulnerability and
Consistency
Watershed models can be classified based on:
a) Nature of the algorithms,
b) Nature of input and uncertainty,
c) Nature of spatial representation etc.
Module 2
Highlights in the Module Contd…
Hydrologic Analysis
Prof. Subhankar KarmakarIIT Bombay
Module 36 Lectures
Module 3
Objective of this module is to introduce the watershedconcepts, rainfall-runoff, hydrograph analysis and unithydrograph theory.
Topics to be covered
Watershed concepts
Characteristics of watershed
Watershed management
Rainfall-runoff
Rational Method
Hydrograph analysis
Hydrograph relations
Recession and Base flow separation
Net storm rainfall and the hydrograph
Time- Area method
Module 3
Topics to be covered
Unit hydrograph theory
Derivation of UH : Gauged watershed
S-curve method
Discrete convolution equation
Synthetic unit hydrograph
Snyder’s method
SCS method
Module 3
(contd..)
Lecture 1: Watershed and rainfall-runoff relationship
Module 3
Watershed concepts
The watershed is the basic unit used in
most hydrologic calculations relating to
water balance or computation of
rainfall-runoff
The watershed boundary (Divide)
defines a contiguous area, such that
the net rainfall or runoff over that area
will contribute to the outlet
The rainfall that falls outside the
watershed boundary will not contribute
to runoff at the outlet
Watershed diagram
Watershed boundary
Module 3
Watershed concepts Contd…
Watersheds are characterized ingeneral, by one main channel andby tributaries that drain into mainchannel at one or more confluencepoints
A “divide” or “drainage divide” is theline drawn through the highestelevated points within a watershed
Divide forms the limits of a singlewatershed and the boundarybetween two or more watersheds
River
stream
Divide
Sub-catchment or Sub-basin
Module 3
• A water divide is categorized into:-1. Surface water divide –highest elevation line between basins
(watersheds) that defines the perimeter and sheds water into adjacentbasins, and,
2. Subsurface water divide –which refers to faults, folds, tilted geologicstrata (rock layers), etc., that cause sub-surface flow to move in onedirection or the other.
Surface water divideSubsurface water divide
Module 3
Watershed concepts Contd…
Size: It helps in computing parameters like rainfall
received, retained, amount of runoff etc.
Shape: Based on the morphological parameters such as geological
structure eg. peer or elongated
Slope: Reflects the rate of change of elevation with distance along the
main channel and controls the rainfall distribution and movement
Drainage: Determines the flow characteristics and the erosion behavior
Soil type: Determines the infiltration rates that can occur for the area
Characteristics of watershed
Module 3
Land use and land cover: It can affect the overland flow of therainwater with the improve in urbanization and increased pavements.
Main channel and tributary characteristics: It can effect the streamflow response in various ways such as slope, cross-sectionalarea, Manning’s roughness coefficient, presence of obstructions andchannel condition
Physiography: Lands altitude and physical disposition
Socio-economics: Depends on the standard of living of the people andit is important in managing water
Module 3
Characteristics of watershed Contd…
A watershed management approach is one that considers the watershed as awhole, rather than separate parts of the watershed in isolation
Managing the water and other natural resources is an effective and efficientway to sustain the local economy and environmental health
Watershed management helps reduce flood damage, decrease the loss ofgreen space, reduce soil erosion and improve water quality
Watershed planning brings together the people within the watershed,regardless of political boundaries, to address a wide array of resourcemanagement issues
Module 3
Watershed Management
Use an ecological approach that would recover and maintain the biologicaldiversity, ecological Function, and defining characteristics of naturalecosystems
Recognize that humans are part of ecosystems-they shape and are shapedby the natural systems: the sustainability of ecological and societal systemsare mutually dependent
Adopt a management approach that recognizes ecosystems and institutionsare characteristically heterogeneous in time and space
Integrate sustained economic and community activity into the managementof ecosystems
Module 3
Principles for Watershed Management
Provide for ecosystem governance at appropriate ecological and institutionalscales
Use adaptive management as the mechanism for achieving both desiredoutcomes and new understandings regarding ecosystem conditions
Integrate the best science available into the decision-making process, whilecontinuing scientific research to reduce uncertainties
Implement ecosystem management principles through coordinatedgovernment and non-government plans and activities
Develop a shared vision of desired human and environmental conditions
Module 3
Principles for Watershed Management Contd…
Lecture 2: Watershed and rainfall-runoff relationship (contd.)
Module 3
Rainfall-Runoff
How does runoff occur?
When rainfall exceeds the infiltration rate at the surface, excess water
begins to accumulate as surface storage in small depressions. As
depression storage begins to fill, overland flow or sheet flow may begin to
occur and this flow is called as “Surface runoff”
Runoff mainly depends on: Amount of rainfall, soil type, evaporation
capacity and land use
Amount of rainfall: The runoff is in direct proportion with the rainfall. i.e.
as the rainfall increases, the chance of increase in runoff will also
increases
Module 3
Soil type: Infiltration rate depends mainly on the soil type. If the soil is having
more void space (porosity), than the infiltration rate will be more causing less
surface runoff (eg. Laterite soil)
Evaporation capacity: If the evaporation capacity is more, surface runoff will
be reduced
Components of Runoff
Overland Flow or Surface Runoff: The water that travels over the ground
surface to a channel. The amount of surface runoff flow may be small since it
may only occur over a permeable soil surface when the rainfall rate exceeds
the local infiltration capacity.
Rainfall-Runoff Contd….
Module 3
Interflow or Subsurface Storm Flow: The precipitation that infiltrates the soilsurface and move laterally through the upper soil layers until it enters a streamchannel.
Groundwater Flow or Base Flow: The portion of precipitation that percolatesdownward until it reaches the water table. This water accretion may eventuallydischarge into the streams if the water table intersects the stream channels ofthe basin. However, its contribution to stream flow cannot fluctuate rapidlybecause of its very low flow velocity
Data collection The local flood control agencies are responsible for extensive hydrologic
gaging networks within India, and data gathered on an hourly or daily basiscan be plotted for a given watershed to relate rainfall to direct runoff for agiven year.
Module 3
Rainfall-Runoff Contd….
• Travel time for open channel flow (Tt)Tt = L/V
where L = length of open channel (ft, m)V = cross-sectional average velocity of flow (ft/s, m/s)
Manning's equation can be used to calculate cross-sectional average velocity of flow in open channels
where V = cross-sectional average velocity (ft/s, m/s)kn = 1.486 for English units and kn = 1.0 for SI unitsA = cross sectional area of flow (ft2, m2)n = Manning coefficient of roughness
R = hydraulic radius (ft, m)S = slope of pipe (ft/ft, m/m)
Module 3
Rational Method
Runoff Measurement Contd….
V = kn / n R2/3 S1/2
Hydraulic radius (R) can be expressed asR = A/P
where A = cross sectional area of flow (ft2,m2)P = wetted perimeter (ft, m)
After getting the value of Tt, the time of concentration can be obtained byTc = ∑Tt
Rational Method
Values of Runoff coefficients, C (Chow, 1962)
Module 3
Runoff Measurement Contd….
Rational Method
Calculation of Tc
• Tc = ∑Tt
where Tt is the travel time i.e. the time it takes for water to travel fromone location to another in a watershed
• Travel time for sheet flow
where,n = Manning’s roughness coefficientL = Flow length (meters)
P2 = 2-yr, 24-hr rainfall (in.) and S is the hydraulic grade line orland surface
Module 3
Rational Method
Runoff Measurement Contd….
• Travel time for open channel flow
• Where V is the velocity of flow (in./hr)
• Hence Tt = L/V
• After getting the value of Tt, the time of concentration can be obtained by
Tc = ∑Tt
Module 3
Rational Method
Runoff Measurement Contd….
• Assumptions of rational method
Steady flow and uniform rainfall rate will produce maximum runoff when all
parts of a watershed are contributing to outflow
Runoff is assumed to reach a maximum when the rainfall intensity lasts as
long as tc
Runoff coefficient is assumed constant during a storm event
• Drawbacks of rational method
The rational method is often used in small urban areas to design drainage
systems and open channels
For larger watersheds, this process is not suitable since this method is
usually limited to basins less than a few hundred acres in size
Module 3
Rational Method
Runoff Measurement Contd….
Lecture 3: Hydrograph analysis
Module 3
Hydrograph analysis
A hydrograph is a continuous plot of instantaneous discharge v/s time. It
results from a combination of physiographic and meteorological conditions in
a watershed and represents the integrated effects of climate, hydrologic
losses, surface runoff, interflow, and ground water flow
Detailed analysis of hydrographs is usually important in flood damage
mitigation, flood forecasting, or establishing design flows for structures that
convey floodwaters
Factors that influence the hydrograph shape and volume
Meteorological factors
Physiographic or watershed factors and
Human factors
Module 3
• Meteorological factors include
Rainfall intensity and pattern
Areal distribution or rainfall over the basin and
Size and duration of the storm event
• Physiographic or watershed factors include
Size and shape of the drainage area
Slope of the land surface and main channel
Channel morphology and drainage type
Soil types and distribution
Storage detention in the watershed
• Human factors include the effects of land use and land cover
Hydrograph analysis Contd…
• During the rainfall, hydrologiclosses such as infiltration,depression storage and detentionstorage must be satisfied prior tothe onset of surface runoff
• As the depth of surface detentionincreases, overland flow may occurin portion if a basin
• Water eventually moves into smallrivulets, small channels and finallythe main stream of a watershed
• Some of the water that infiltratesthe soil may move laterally throughupper soil zones (subsurfacestromflow) until it enters a streamchannel
Uniform rainfall
Infiltration
Depression storage
Detention storage
Time (hr)
Run
off (
cfs)
Rai
nfal
l (in
./hr)
Distribution of uniform rainfall
Hydrograph analysis Contd…
• If the rainfall continues at a constantintensity for a very long period,storage is filled at some point andthen an equilibrium discharge canbe reached
• In equilibrium discharge the inflow andoutflow are equal
• The point P indicates the time atwhich the entire discharge areacontributes to the flow
• The condition of equilibrium dischargeis seldom observed in nature, exceptfor very small basins, because ofnatural variations in rainfall intensityand duration
Rainfall
Equilibrium discharge
Run
off (
cfs)
Rai
nfal
l (in
./hr)
Time (hr)
P
Equilibrium hydrograph
Module 3
Hydrograph analysis Contd…
Hydrograph relations
• The typical hydrograph is
characterized by a
1. Rising limb
2. Crest
3. Recession curve
• The inflation point on the falling limb
is often assumed to be the point
where direct runoff ends
Net rainfall = Vol. DRO
Crest
Falling limb
Inflation point
Recession
Direct runoff(DRO)
Recession
Rising limb
Pn
Q
Time
Base flow (BF)
Hydrograph relations
Module 3
Hydrograph analysis Contd…
Recession and Base flow separation• In this the hydrograph is divided into
two parts1. Direct runoff (DRO) and2. Base flow (BF)
• DRO include some interflow whereasBF is considered to be mostly fromcontributing ground water
• Recession curve method is used toseparate DRO from BF and can by anexponential depletion equation
qt = qo ·e-kt whereqt = discharge at a later time tqo = specified initial dischargek = recession constant
C
D
BA
Q
Time
N=bA0.2
Base flow separationModule 3
Hydrograph analysis Contd…
• There are three types of baseflow separation techniques 1. Straight line method2. Fixed base method3. Constant slope method
1. Straight line method• Assume baseflow constant regardless of stream height (discharge)• Draw a horizontal line segment (A-D) from beginning of runoff to intersection
with recession curve2. Constant slope method• connect inflection point on receding limb of storm hydrograph to beginning of
storm hydrograph• Assumes flow from aquifers began prior to start of current storm, arbitrarily
sets it to inflection point• Draw a line connecting the point (A-C) connecting a point N time periods
after the peak.
Module 3
Baseflow Separation Methods
3. Fixed Base Method
• Assume baseflow decreases while stream flow increases (i.e. to peak of
storm hydrograph)
• Draw line segment (A –B) from baseflow recession to a point directly below
the hydrograph peak
• Draw line segment (B-C) connecting a point N time periods after the peak
where
N = time in days where DRO is terminated, A= Discharge area in km2,
b= coefficient, taken as 0.827
Module 3
Baseflow Separation Methods Contd…
The distribution of gross rainfall can be given by the continuity equation as
Gross rainfall = depression storage+ evaporation+ infiltration+surface runoff
In case, where depression storage is small and evaporation can beneglected, we can compute rainfall excess which equals to direct runoff,DRO, by
Rainfall excess (Pn) = DRO = gross rainfall – (infiltration+depression storage)
Module 3
Rainfall excess
• The simpler method to determinerainfall excess include1. Horton infiltration method2. Ø index method
• Note:- In this, the initial loss isincluded for depression storage
Rai
nfal
l and
infil
tratio
n
Depression storage
Net storm rainfall
Ø index
Horton infiltration
Time
Infiltration loss curves
Module 3
Rainfall excess Contd…
• Horton infiltration methodHorton method estimates infiltration with an exponential-type equation that
slowly declines in time as rainfall continues and is given by
f= fc + (fo – fc) e-kt ( when rainfall intensity i>f)where
f = infiltration capacity (in./hr)fo = initial infiltration capacity (in./hr)fc = final infiltration capacity (in./hr)k = empirical constant (hr-1)
• Ø index methodIt is the simplest method and is calculated by finding the loss difference
between gross precipitation and observed surface runoff measured as a hydrograph
Module 3
Rainfall excess Contd…
• Rainfall of magnitude 3.8 cm and 2.8 cm occurring on two consecutive 4-hdurations on a catchment area 27km2 produced the following hydrograph offlow at the outlet of the catchment. Estimate the rainfall excess and φ-index
Time from start of rainfall (h) -6 0 6 12 18 24 30 36 42 48 54 60 66Observed flow (m3/s) 6 5 13 26 21 16 12 9 7 5 5 4.5 4.5
Example Problem-1
Module 3
Baseflow separation:
Using Simple straight line method,
N = 0.83 A0.2 = 0.83 (27)0.2
= 1.6 days = 38.5 h
So the baseflow starts at 0th h and ends at the point (12+38.5)h
Hydrograph
6 5
13
26
21
16
129
75 5 4.5 4.5
0
5
10
15
20
25
30
-10 -5 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70
Dis
char
ge(m
3 /s)
Time (hr)
Module 3
50.5 h ( say 48 h approx.)
Constant baseflow of 5m3/s
Example Problem-1 Contd…
Time (h) FH Ordinates(m3/s) DRH Ordinates (m3/s)-6 6 10 5 06 13 8
12 26 2118 21 1624 16 1130 12 736 9 442 7 248 5 054 5 060 4.5 066 4.5 0
DRH ordinates are obtained from subtracting the corresponding FH with the base flow i.e. 5 m3/s Module 3
Example Problem-1 Contd…
Hydrograph
6 5
13
26
21
16
12
97
5 5 4.5 4.5
0
5
10
15
20
25
30
-10 -5 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70
Dis
char
ge(m
3 /s)
Time (hr)
Area of Direct runoff hydrograph
Module 3
Example Problem-1 Contd…
Area of DRH = (6*60*60)[1/2 (8)+1/2 (8+21)+
1/2 (21+16)+ 1/2 (16+11)+
1/2 (11+7)+ 1/2 (7+4)+ 1/2 (4+2)+ 1/2 (2)]
= 1.4904 * 106m3 (total direct runoff due to storm)
Run-off depth = Runoff volume/catchment area
= 1.4904 * 106/27* 106
= 0.0552m = 5.52 cm = rainfall excess
Total rainfall = 3.8 +2.8 = 6.6cm
Duration = 8h
φ-index = (P-R)/t = (6.6-5.52)/8 = 0.135cm/h
Module 3
Example Problem-1 Contd…
A storm over a catchment of area 5.0 km2 had a duration of 14hours. The masscurve of rainfall of the storm is as follows:
If the φ-index of the catchment is 0.4cm/h, determine the effective rainfallhyetograph and the volume of direct runoff from the catchment due to thestorm.
Time from start of
storm (h) 0 2 4 6 8 10 12 14Accumulated rainfall
(cm) 0 0.6 2.8 5.2 6.6 7.5 9.2 9.6
Module 3
Example Problem-2
Time from start of
storm(h)Time
interval ∆t
Accumulated rainfall in ∆t
(cm)
Depth of rainfall in ∆t (cm)
φ ∆t (cm) ER (cm)
Intensity of ER (cm/h)
0 _ 0 _ _ _ _2 2 0.6 0.6 0.8 0 04 2 2.8 2.2 0.8 1.4 0.76 2 5.2 2.4 0.8 1.6 0.88 2 6.7 1.5 0.8 0.7 0.35
10 2 7.5 0.8 0.8 0 012 2 9.2 1.7 0.8 0.9 0.4514 2 9.6 0.4 0.8 0 0
Module 3
Example Problem-2 Contd…
• Total effective rainfall = Direct runoff due to storm = area of ER hyetograph= (0.7+0.8+0.35+0.45)*2 = 4.6 cm
• Volume of direct runoff = (4.6/100) * 5.0*(1000)2
= 230000m3
Run-off Measurement
• This method assumes that the outflow hydrograph results from puretranslation of direct runoff to the outlet, at an uniform velocity, ignoring anystorage effect in the watershed
• The relation ship is defined by dividing a watershed into subareas withdistinct runoff translation times to the outlet
• The subareas are delineated with isochrones of equal translation timenumbered upstream from the outlet
• In a uniform rainfall intensity distribution over the watershed, water first flowsfrom areas immediately adjacent to the outlet, and the percentage of totalarea contributing increases progressively in time
• The surface runoff from area A1 reaches the outlet first followed bycontributions from A2, A3 and A4,
Module 3
Time- Area method
2A1A
3A4A
Isochrone of Equal time to outlet
hr5hr10hr15
jiin ARARARQ 1211 ...+++= −
2R
1R3R
Time, t
Rai
nfal
l
2A
1A
3A
4A
0 5 10 15 20Time, t
Are
a
Outlet
Module 3
Run-off Measurement Contd…
Time- Area method
whereQn = hydrograph ordinate at time n (cfs)Ri = excess rainfall ordinate at time i (cfs)Aj = time –area histogram ordinate at time j (ft2)
Limitation of time area method
• This method is limited because of the difficulty of constructing isochronallines and the hydrograph must be further adjusted to represent storageeffects in the watershed
Module 3
Time- Area method
Run-off Measurement Contd…
• Find the storm hydrograph for the following data using time area method.Given rainfall excess ordinate at time is 0.5 in./hr
A B C DArea (ac) 100 200 300 100Time to gage G (hr) 1 2 3 4
AB
CD
G
Module 3
Time area histogram method uses Qn = RiA1 + Ri-2A2 +…….+ RiAj
For n = 5, i = 5, and j = 5Q5 = R5A1 + R4A2+ R3A3 + R2A4
(0.5 in./ hr) (100 ac) + (0.5 in./hr) (200 ac) + (0.5 in./hr) (300ac) + (0.5 in./hr) (100)
Q5 = 350 ac-in./hrNote that 1 ac-in./hr ≈ 1 cfs, hence
Q5 = 350 cfs
Example Problem
Example Problem Contd…
Time(hr)
HydrographOrdinate (R1:Rn)
Basin No.
Time to gage
Basin area A1:An (ac)
R1:An R2:An R2:An R2:An R2:An Stormhydrograph
0 0
1 0.5 A 1 100 * 50 50
2 0.5 B 2 200 100 50 +150
3 0.5 C 3 300 150 100 50 300
4 0.5 D 4 400 50 150 100 50 350
5 50 150 100 50 350
6 50 150 100 300
7 50 150 200
8 50 50
9 0
Excel spreadsheet calculation
* =(R1*A1) = (0.5*100) and + = (adding the columns from 6 to 10) Module 3
0
50
100
150
200
250
300
350
400
0 1 2 3 4 5 6 7 8 9 10
Contribution of each sub area
A A A A
B B B
C C
D
Time (hr)
Q (C
FS)
Module 3
Example Problem Contd…
Lecture 4: Introduction to unit hydrograph
Module 3
Unit hydrograph (UH)
• The unit hydrograph is the unit pulse response function of a linear hydrologicsystem.
• First proposed by Sherman (1932), the unit hydrograph (originally namedunit-graph) of a watershed is defined as a direct runoff hydrograph (DRH)resulting from 1 in (usually taken as 1 cm in SI units) of excess rainfallgenerated uniformly over the drainage area at a constant rate for an effectiveduration.
• Sherman originally used the word “unit” to denote a unit of time. But sincethat time it has often been interpreted as a unit depth of excess rainfall.
• Sherman classified runoff into surface runoff and groundwater runoff anddefined the unit hydrograph for use only with surface runoff.
Module 3
The unit hydrograph is a simple linear model that can be used to derive the
hydrograph resulting from any amount of excess rainfall. The following basic
assumptions are inherent in this model;
1. Rainfall excess of equal duration are assumed to produce hydrographs
with equivalent time bases regardless of the intensity of the rain
2. Direct runoff ordinates for a storm of given duration are assumed directly
proportional to rainfall excess volumes.
3. The time distribution of direct runoff is assumed independent of
antecedent precipitation
4. Rainfall distribution is assumed to be the same for all storms of equal
duration, both spatially and temporally
Unit hydrograph Contd….
Module 3
Terminologies1. Duration of effective rainfall : the time
from start to finish of effective rainfall
2. Lag time (L or tp): the time from the
center of mass of rainfall excess to the
peak of the hydrograph
3. Time of rise (TR): the time from the start
of rainfall excess to the peak of the
hydrograph
4. Time base (Tb): the total duration of the
DRO hydrographBase flow
Direct runoff
Inflection point
TRtp
Effective rainfall/excess rainfall
Q (c
fs)
Module 3
Derivation of UH : Gauged watershed
1. Storms should be selected with a simple structure with relatively uniform spatial
and temporal distributions
2. Watershed sizes should generally fall between 1.0 and 100 mi2 in modern
watershed analysis
3. Direct runoff should range 0.5 to 2 in.
4. Duration of rainfall excess D should be approximately 25% to 30% of lag time tp
5. A number of storms of similar duration should be analyzed to obtain an average
UH for that duration
6. Step 5 should be repeated for several rainfall of different durations
Module 3
Unit hydrograph
Rules to be observed in developing UH from gaged watersheds
1. Analyze the hydrograph and separate base flow
2. Measure the total volume of DRO under the hydrograph and convert time to
inches (mm) over the watershed
3. Convert total rainfall to rainfall excess through infiltration methods, such that
rainfall excess = DRO, and evaluate duration D of the rainfall excess that
produced the DRO hydrograph
4. Divide the ordinates of the DRO hydrograph by the volume in inches (mm)
and plot these results as the UH for the basin. Time base Tb is assumed
constant for storms of equal duration and thus it will not change
5. Check the volume of the UH to make sure it is 1.0 in.(1.0mm), and graphically
adjust ordinates as required
Module 3
Unit hydrograph
Essential steps for developing UH from single storm hydrograph
Obtain a Unit Hydrograph for a basin of 315 km2 of area using the rainfall and stream flow data tabulated below.
Time (hr) Observed hydrograph(m3/s)
0 100
1 100
2 300
3 700
4 1000
5 800
6 600
7 400
8 300
9 200
10 100
11 100
Time (hr)
Gross PPT(GRH) (cm/h)
0-1 0.51-2 2.52-3 2.53-4 0.5
Stream flow data Rainfall data
Module 3
Unit hydrograph
Example Problem
• Empirical unit hydrograph derivation separates the base flow from the observedstream flow hydrograph in order to obtain the direct runoff hydrograph (DRH). Forthis example, use the horizontal line method to separate the base flow. Fromobservation of the hydrograph data, the stream flow at the start of the rising limbof the hydrograph is 100 m3/s
• Compute the volume of direct runoff. This volume must be equal to the volume of the effective rainfall hyetograph (ERH)
VDRH = (200+600+900+700+500+300+200+100) m3/s (3600) s = 12'600,000 m3
• Express VDRH in equivalent units of depth:
VDRH in equivalent units of depth = VDRH/Abasin = 12'600,000 m3/(315000000 m2) = 0.04 m = 4 cm
Module 3
Unit hydrograph
Example Problem Contd…
Obtain a Unit Hydrograph by normalizing the DRH. Normalizing implies dividing theordinates of the DRH by the VDRH in equivalent units of depth
Time (hr) Observed hydrograph(m3/s)
Direct Runoff Hydrograph (DRH) (m3/s)
Unit Hydrograph (m3/s/cm)
0 100 0 01 100 0 02 300 200 503 700 600 1504 1000 900 2255 800 700 1756 600 500 1257 400 300 758 300 200 509 200 100 25
10 100 0 011 100 0 0
Module 3
Module 3
Unit hydrograph
Example Problem Contd…
0
200
400
600
800
1000
1200
0 2 4 6 8 10 12
Q (m
3 /s)
Time (hr)
Observed hydrograph
Unit hydrograph
DRH
• Determine the duration D of the ERH associated with the UH obtained in 4. In order to do this:
1. Determine the volume of losses, VLosses which is equal to the difference between the volume of gross rainfall, VGRH, and the volume of the direct runoff hydrograph, VDRH . VLosses = VGRH - VDRH = (0.5 + 2.5 + 2.5 +0.5) cm/h 1 h - 4 cm = 2 cm
2. Compute the f-index equal to the ratio of the volume of losses to the rainfall duration, tr. Thus, ø-index = VLosses/tr = 2 cm / 4 h = 0.5 cm/h
3. Determine the ERH by subtracting the infiltration (e.g., ø-index) from the GRH:
Module 3
Unit hydrograph
Example Problem Contd…
Time (hr) Effective precipitation (ERH)
(cm/hr)0-1 0
1-2 2
2-3 2
3-4 0
As observed in the table, the duration of the effective rainfall hyetograph is 2 hours.Thus, D = 2 hours, and the Unit Hydrograph obtained above is a 2-hour UnitHydrograph.
Module 3
Unit hydrograph
Example Problem Contd…
Lecture 5: Derivation of S-curve and discrete convolution equations
Module 3
• It is the hydrograph of direct surface discharge that would result from a
continuous succession of unit storms producing 1cm(in.)in tr –hr
• If the time base of the unit hydrograph is Tb hr, it reaches constant outflow
(Qe) at T hr, since 1 cm of net rain on the catchment is being supplied and
removed every tr hour and only T/tr unit graphs are necessary to produce an
S-curve and develop constant outflow given by,
Qe = (2.78·A) / tr
where
Qe = constant outflow (cumec)
tr = duration of the unit graph (hr)
A = area of the basin (km2 or acres)Module 3
Unit hydrograph
S – Curve method
Unit hydrograph inSuccession produce Constant outflow Qe cumec
Time t (hr)trI
Lagged s-curveLagged by tr-hr
S-curvehydrograph
To obtain tr-hr UG multiply the S-curve difference by tr/trI
Constant flow Qe (Cumec)
Successive unit storms of Pnet = 1 cm
Dis
char
ge Q
(Cum
ec)
Inte
nsity
(cm
/hr)
Lagged
Changing the duration of UG by S-curve technique
Module 3
Unit hydrograph
S – Curve method Contd…
• Convert the following 2-hr UH to a 3-hr UH using the S-curve method
Time (hr) 2-hr UH ordinate (cfs)0 01 752 2503 3004 2755 2006 1007 758 509 2510 0
Module 3
Example Problem
Unit hydrograph
SolutionMake a spreadsheet with the 2-hr
UH ordinates, then copy them in
the next column lagged by D=2
hours. Keep adding columns until
the row sums are fairly constant.
The sums are the ordinates of your
S-curve
Module 3
Unit hydrograph
Time (hr)
2-hr UH
2-HR lagged UH’s Sum
0 0 01 75 752 250 0 2503 300 75 3754 275 250 0 5255 200 300 75 5756 100 275 250 0 6257 75 200 300 75 6508 50 100 275 250 0 6759 25 75 200 300 75 675
10 0 50 100 275 250 0 67511 25 75 200 300 75 675
Example Problem Contd…
0
100
200
300
400
500
600
700
800
0 2 4 6 8 10 12 14
Q (c
fs)
Time (hr)
S-curve
2 hr UHLagged by 2 hr
Draw your S-curve, as shown in figure below
Make a spreadsheet with the 2-hr UH ordinates, then copy them in the next column lagged by D=2 hours. Keep adding columns until the row sums are fairly constant. The sums are the ordinates of your S-curve.
Module 3
Unit hydrograph
Example Problem Contd…
Time (hr) S-curveordinate
S-curve lagged 3hr
Difference 3-HR UH ordinate
0 0 0 01 75 75 502 250 250 166.73 375 0 375 2504 525 75 450 3005 575 250 352 2166 625 375 250 166.77 650 525 125 83.38 675 575 100 66.79 675 625 50 33.3
10 675 650 25 16.711 675 675 0 0
Unit hydrograph
Example Problem Contd…
Find the one hour unit hydrograph using the excess rainfall hyetograph anddirect runoff hydrograph given in the table
Time (1hr) Excess Rainfall (in) Direct Runoff (cfs)
1 1.06 428
2 1.93 1923
3 1.81 5297
4 9131
5 10625
6 7834
7 3921
8 1846
9 1402
10 830
11 313
Module 3
Example Problem
Unit hydrograph
Unit hydrograph Contd….
Discrete Convolution Equation
∑ − +=
=m*
n m n m 1m 1
Q P U m* = min(n,M)
Where Qn = Direct runoff
Pm = Excess rainfall
Un-m+1 = Unit hydrograph ordinates
Suppose that there are M pulses of excess rainfall.
If N pulses of direct runoff are considered, then N equations can be written Qn in terms of N-M+1unknown values of unit hydrograph ordinates, where n= 1, 2, …,N.
Unit hydrograph Contd….
P1 P2 P3
Input Pn
U1U2 U3 U4 U5
Un-m+1
n-m+1
Unit pulse response applied to P1
Unit pulse response applied to P2
n-m+1Un-m+1
Output Qn
Output ∑ − +=
=m*
n m n m 1m 1
Q P U
Combination of 3 rainfall UH
The set of equations for discrete time convolution
∑ − +=
=m*
n m n m 1m 1
Q P U
n = 1, 2,…,N
=1 1 1Q PU
= +2 2 1 1 2Q P U PU
= + +3 3 1 2 2 1 3Q P U P U PU
−= + + +M M 1 M 1 2 1 MQ P U P U ..... PU
+ += + + + +M 1 M 2 2 M 1 M 1Q 0 P U ..... P U PU
− − − − += + + + + + + +N 1 M N M M 1 N M 1Q 0 0 ..... 0 0 ..... P U P U
− − += + + + + + + +N M 1 N M 1Q 0 0 ..... 0 0 ..... 0 P U
Unit hydrograph Contd….
Solution• The ERH and DRH in table have M=3 and N=11 pulses respectively.
• Hence, the number of pulses in the unit hydrograph is N-M+1=11-3+1=9.
• Substituting the ordinates of the ERH and DRH into the equations in tableyields a set of 11 simultaneous equations
Module 3
Unit hydrograph
− −= = =2 2 1
11
Q P U 1,928 1.93x404U 1,079 cfs/ inP 1.06
Similarly calculate for remaining ordinates and the final UH is tabulated below
n 1 2 3 4 5 6 7 8 9Un (cfs/in) 404 1,079 2,343 2,506 1,460 453 381 274 173
Example Problem Contd…
Lecture 6: Synthetic unit hydrograph
Module 3
Synthetic Unit Hydrograph
• In India, only a small number of streams are gauged (i.e., stream flows dueto single and multiple storms, are measured)
• There are many drainage basins (catchments) for which no stream flowrecords are available and unit hydrographs may be required for such basins
• In such cases, hydrographs may be synthesized directly from othercatchments, which are hydrologically and meteorologically homogeneous,or indirectly from other catchments through the application of empiricalrelationship
• Methods for synthesizing hydrographs for ungauged areas have beendeveloped from time to time by Bernard, Clark, McCarthy and Snyder. Thebest known approach is due to Snyder (1938)
Module 3
• Snyder (1938) was the to develop a synthetic UH based on a study of
watersheds in the Appalachian Highlands. In basins ranging from 10 –
10,000 mi.2
Snyder relations are
tp = Ct(LLC)0.3
where
tp= basin lag (hr)
L= length of the main stream from the outlet to the divide (mi)
Lc = length along the main stream to a point nearest the watershed
centroid (mi)
Ct= Coefficient usually ranging from 1.8 to 2.2
Module 3
Snyder’s method
Synthetic unit hydrograph
Qp = 640 CpA/tpwhereQp = peak discharge of the UH (cfs)A = Drainage area (mi2)Cp = storage coefficient ranging from 0.4 to
0.8, where larger values of cp are associated with smaller values of Ct
Tb = 3+tp/8where Tb is the time base of hydrographNote: For small watershed the above eq. should be replaced by multiplying tp by the value varies from 3-5
• The above 3 equations define points for a UH produced by an excess rainfall of duration D= tp/5.5 Snyder’s hydrograph parameter
Snyder’s method Contd…
Module 3
Synthetic unit hydrograph
Use Snyder’s method to develop a UH for the area of 100mi2 described below.Sketch the appropriate shape. What duration rainfall does this correspond to?
Ct = 1.8, L= 18mi,Cp = 0.6, Lc= 10mi
Calculate tp
tp = Ct(LLC)0.3
= 1.8(18·10) 0.3 hr,= 8.6 hr
Module 3
Example Problem
Synthetic unit hydrograph
Calculate Qp
Qp= 640(cp)(A)/tp= 640(0.6)(100)/8.6
= 4465 cfs
Since this is a small watershed,Tb ≈4tp = 4(8.6)
= 34.4 hr
Duration of rainfallD= tp/5.5 hr
= 8.6/5.5 hr= 1.6 hr
0
1000
2000
3000
4000
5000
0 5 10 15 20 25 30 35 40
Q (c
fs)
Time (hr)
Qp
W75
W50 Area drawn to represent 1in. of runoff over thewatershed
W75 = 440(QP/A)-1.08
W50 = 770(QP/A)-1.08
(widths are distributed 1/3 before Qpand 2/3 after)
Module 3
Synthetic unit hydrograph
Example Problem Contd…
• Unit = 1 inch of runoff (not rainfall) in 1
hour
• Can be scaled to other depths and times
• Based on unit hydrographs from many
watersheds
• The earliest method assumed a
hydrograph as a simple triangle, with
rainfall duration D, time of rise TR (hr),
time of fall B. and peak flow Qp (cfs).
tp
Qp
TR B
SCS triangular UH
Module 3
SCS (Soil Conservation Service) Unit Hydrograph
Synthetic unit hydrograph
• The volume of direct runoff is
orwhere B is given by
Therefore runoff eq. becomes, for 1 in. of rainfall excess,
=
BTvolQ
Rp +=
2
RTB 67.1=
Rp T
volQ 75.0=
Rp T
AQ 484=
whereA= area of basin (sq mi)TR = time of rise (hr)
Module 3
Rp T
AQ )008.1()640(75.0=
22BQTQ
Vol pRp +=
SCS Unit Hydrograph Contd…
Synthetic unit hydrograph
• Time of rise TR is given by
whereD= rainfall duration (hr)tp= lag time from centroid of rainfall to QP
Lag time is given by
whereL= length to divide (ft)Y= average watershed slope (in present)CN= curve number for various soil/land use
Module 3
pR tDT +=2
0.5
0.70.8
L
19000y
9CN
1000
−
=pt
SCS Unit Hydrograph Contd…
Synthetic unit hydrograph
Runoff curve number for different land use (source: Woo-Sung et al.,1998)
Module 3
SCS Unit Hydrograph Contd…
Synthetic unit hydrograph
Use the SCS method to develop a UH for the area of 10 mi2 described below.Use rainfall duration of D = 2 hr
Ct = 1.8, L= 5mi,Cp = 0.6, Lc= 2mi
The watershed consist CN = 78 and the average slope in the watershed is 100 ft/mi. Sketch the resulting SCS triangular hydrograph .
Module 3
Example Problem
Synthetic unit hydrograph
SolutionFind tp by the eq.
Convert L= 5mi, or (5*5280 ft/mi) = 26400 ft.
Slope is 100 ft/mi, so y = (100ft/mi) (1mi/5280 ft)(100%) = 1.9%
Substituting these values in eq. of tp, we get tp = 3.36 hr
0.5
0.70.8
L
19000y
9CN
1000
−
=pt
Find TR using eq.
Given rainfall duration is 2 hr, TR = 4.36 hr, the rise of the hydrographThen find Qp using the eq, given A= 10 mi2
. Hence Qp = 1.110 cfs
Module 3
Synthetic unit hydrograph
Rp T
AQ 484=
pR tDT +=2
To complete the graph, it is also necessary to know the time of fall B. The
volume is known to be 1 in. of direct runoff over the watershed.
So, Vol. = (10mi2) (5280ft/mi)2 (ac/43560ft2) (1 in.) = 6400 ac-in
Hence from eq.
B = 7.17 hr22BQTQ
Vol pRp +=
Example Problem Contd…
0
200
400
600
800
1000
1200
0 2 4 6 8 10 12 14
Q (c
fs)
Time (hr)
Qp= 1110 (cfs)
TR=4.36 (hr) B=7.17 (hr)
Module 3
Synthetic unit hydrograph
Example Problem Contd…
Exercise problems
1. The stream flows due to three successive storms of 2.9, 4.9 and 3.9 cm of 6hours duration each on a basin are given below. The area of the basin is 118.8km2 . Assuming a constant base flow of 20 cumec, derive a 6-hour unithydrograph for the basin. An average storm loss of 0.15 cm/hr can be assumed(Hint :- Use UH convolution method)
Time (hr) 0 3 6 9 12 15 18 21 24 27 30 33
Flow (cumec)
20 50 92 140 199 202 204 144 84 45 29 20
Module 3
2. The ordinates of a 4-hour unit hydrograph for a particular basin are givenbelow. Derive the ordinates of (i) the S-curve hydrograph, and (ii) the 2-hourunit hydrograph, and plot them, area of the basin is 630 km2
Time (hr) Discharge (cumec)0 02 254 1006 1608 190
10 17012 110
Time (hr) Discharge (cumec)14 7016 3018 2020 622 1.524 0
Module 3
Exercise problems Contd…
3. The following are the ordinates of the 9-hour unit hydrograph for the entirecatchment of the river Damodar up to Tenughat dam site: and the catchmentcharacteristics are , A = 4480 km2, L = 318 km, Lca = 198 km. Derive a 3-hourunit hydrograph for the catchment area of river Damodar up to the head ofTenughat reservoir, given the catchment characteristics as, A = 3780km2, L =284 km, Lca = 184km. Use Snyder’s approach with necessary modifications forthe shape of the hydrograph.
Time (hr) 0 9 18 27 36 45 54 63 72 81 90
Flow (cumec)
0 69 1000 210 118 74 46 26 13 4 0
Module 3
Exercise problems Contd…
This module presents the concept of Rainfall-Runoff analysis, or the
conversion of precipitation to runoff or streamflow, which is a central
problem of engineering hydrology.
Gross rainfall must be adjusted for losses to infiltration, evaporation and
depression storage to obtain rainfall excess, which equals Direct Runoff
(DRO).
The concept of the Unit hydrograph allows for the conversion of rainfall
excess into a basin hydrograph, through lagging procedure called
hydrograph convolution.
The concept of synthetic hydrograph allows the construction of hydrograph,
where no streamflow data are available for the particular catchment.
Module 3
Highlights in the Module
Hydrologic Analysis(Contd.)
Prof. Subhankar KarmakarIIT Bombay
Module 43 Lectures
Module 4
Objective of this module is to learn linear-kinematicwave models and overland flow models
Topics to be covered
Kinematic wave modeling Continuity equation
Momentum equation
Saint Venant equation
Kinematic overland flow modeling
Kinematic channel modeling
Module 4
Lecture 1: Kinematic wave method
Module 4
Kinematic wave method
• This method assumes that the weight or gravity force of flowingwater is simply balanced by the resistive forces of bed friction
• This method can be used to derive overland flow hydrographs,which can be added to produce collector or channel hydrographsand eventually, as stream or channel hydrograph
• This method is the combination of continuity equation and asimplified form of St. Venant equations
(Note:- The complete description of St. Venant equations is providedin Module-6)
Module 4
Q
S0
Ѳ
Z
V2/2g
h
dx
Ff
x
dxxQQ∂∂
+
Datum
Continuity Equation
y
FH
Fg
Energy line
Kinematic modeling methods
Module 4
The general equation of continuity,Inflow-Outflow = rate of change of storage
Inflow =
Outflow =
Storage change = txtA
∆∆∂∂
where,q= rate of lateral inflow per unit length of channelA = cross- sectional area
Kinematic modeling methods
Continuity Equation Contd…
Module 4
txqtxxQQ ∆∆+∆•
∆
•∂∂
−2
txxQQ ƥ
∆
•∂∂
−2
The equation of continuity becomes, after dividing by ∆x and ∆t,
• For unit width b of channel with v= average velocity, the continuity equation
can be written as
qxQ
tA
=∂∂
+∂∂
Module 4
bq
ty
xyv
xvy =
∂∂
+∂∂
+∂∂
Kinematic modeling methods
Continuity Equation Contd…
Momentum equation
It is based on Newton’s second law and that is, Net force = rate of change of momentum
The following are the three main external forces are acting on area A
Hydrostatic : FH =
Gravitational : Fg=
Frictional : Ff=
= specific weight of water (ρg)
y= distance from the water surface to the centroid of the pressure prism
Sf= friction slope, obtained by solving for the slope in a uniform flow equation, (manning’s equation)
So= Bed slope
γ
Kinematic modeling methods
Module 4
( ) xxyA
∆∂
∂− γ
AS xγ− ∆0
fAS xγ− ∆
• The rate of change of momentum is expressed from Newton’s second lawas
where the total derivative of v W.R.T t can be expressed
( )mvdtdF =
xvv
tv
dtdv
∂∂
+∂∂
=
………..4.1
………..4.2
Module 4
Momentum Equation Contd…
Kinematic modeling methods
• Equating Eq. 4.1 to the sum of the three external forces results in
= g(So-Sf)
• For negligible lateral inflow and a wide channel, the Eq. 4.3 can berearranged to yield
Sf = So
………..4.3
………..4.4
Saint Venant equation
Module 4
( )Avq
xyA
Ag
xvv
tv
+∂
∂+
∂∂
+∂∂
tgv
xgvv
xy
∂∂
−∂∂
−∂∂
−1
Momentum Equation Contd…
Kinematic modeling methods
• In developing the general unsteady flow equation it is assumed that the flow is
one-dimensional (variation of flow depth and velocity are considered to vary
only in the longitudinal X- direction of the channel
• The velocity is constant and the water surface is horizontal across any section
perpendicular to the longitudinal flow axis
• All flows are gradually varied with hydrostatic pressure such that all the vertical
accelerations within the water column cab be neglected
• The longitudinal axis of the flow channel can be approximated by a straight
line, therefore, no lateral secondary circulations occur
Assumptions of Saint Venant equations
Module 4
Kinematic modeling methods
• The slope of the channel bottom is small (less than 1:10)
• The channel boundaries may be treated as fixed non-eroding and non-aggarading
• Resistance to flow may be described by empirical resistance equations suchas the manning or Chezy equations
• The flow is incompressible and homogeneous in density
Module 4
Assumptions of Saint Venant equations Contd…
Kinematic modeling methods
Forms of momentum Equation
Kinematic modeling methods
Module 4
Type of flow Momentum equationKinematic wave ( study uniform)
Sf = So
Diffusion (non inertia) model Sf = So
Steady no-uniform Sf = So
Unsteady non-uniform Sf = So
xy∂∂
−
Dynamic wave
xv
gv
xy
∂∂
−
∂∂
−
tv
g1
xv
gv
xy
∂∂
−
∂∂
−
∂∂
−
Possible types of open channel flow
Module 4
Kinematic modeling methods
Kinematic wave Dynamic wave
It is defined as the study of motionexclusive of the influences of massand force
In this the influences of mass and forceare included
When the inertial and pressure forcesare not important to the movement ofwave then the kinematic wavesgoverns the flow
When inertial and pressure forces areimportant then dynamic waves governthe moment of long waves in shallowwater (large flood wave in a wide river)
Force of this nature will remainapproximately uniform all along thechannel (Steady and uniform flow)
Flows of this nature will be unsteadyand non-uniform along the length ofthe channel
Froude No. < 2 Froude No. > 2
Difference between kinematic and dynamic wave
Kinematic modelling methods
Fr =
Froude number
gdv
WhereV= velocity of flowg= acceleration due to gravityd= hydraulic depth of water
Wave celerity (C) gdc =1. Flows with Froude numbers greater than one are classified as supercritical
flows2. Froude number greater than two tend to be unstable, that are classified as
dynamic wave3. Froude number less then 2 are classified as kinematic wave
Kinematic modelling methods
Module 4
Visualization of dynamic and kinematic waves
Kinematic modelling methods
Module 4
Lecture 2: Kinematic overland flow routing
Module 4
• For the conditions of kinematic flow, and with no appreciable backwater effect,
the discharge can be described as a function of area only, for all x and t;
Q= α · Am
where,
Q= discharge in cfs
A= cross-sectional area
α , m = kinematic wave routing parameters
Kinematic overland flow routing
………..4.5
Module 4
• Henderson (1966) normalized momentum Eq. 4.4 in the form of
Governing equations
………..4.6
Less than one, than the equation will represent Kinematic flow
where Qo=flow under uniform
condition
Hence, for the kinematic flow condition,
Q≈Qo ………..4.7
Kinematic routing methods
Module 4
21
111
+
∂∂
+∂∂
+∂∂
−=gyqv
tgv
xgvv
xy
SQQ
oo
• Woolhiser and Liggett (1967) analyzed characteristics of the rising overland
flow hydrograph and found that the dynamic terms can generally be
neglected if,
or
where, L= length of the planeFr= Froude numbery= depth at the end of the planeS0= slopek= dimensionless kinematic flow number
………..4.8
Kinematic routing methods
Module 4
102 ≥=yFr
LSk o 102 ≥=
vLgS
k o
Governing equations Contd…
Q* is the dimensionless flow v/s t* (dimensionless time) for varies values of k in Eq. 8. It can be seen that for k≤10, large errors in calculation of Q* result by deleting dynamic terms from the momentum Eq. for overland flow
Effect of kinematic wave number k on the rising hydrograph
Module 4
Kinematic routing methods
Governing equations Contd…
• The momentum Eq. for an overland flow segment on a wide plane withshallow flows can be derived from Eq. 4.5 and manning's Eq. for overlandflow
• Rewriting the Eq. 4.9 in terms of flow per unit width for an overland flow qo,we have
………..4.9
= conveyance factor
mo= 5/3 from manning’s Eq.So= Average overland flow slopeyo= mean depth of overland flow
………..4.10
Module 4
3/5ySn
kq o
m=
omooo yq α= o
mo S
nk
=α
Kinematic routing methods
Governing equations Contd…
Estimates of Manning’s roughness coefficients for overland flow
Kinematic routing methods
Module 4
• The continuity Eq. is
Finally, by substituting Eq. 4.11 in Eq. 4.9, we have
Eq. 4.10 and Eq.4.12 form the complete kinematic wave equation for overland flow
where,i= rate of gross rainfall (ft/s)f= infiltration rateqo= flow per unit width ( cfs/ft)yo= mean depth of overland flow
………..4.11
………..4.12
Module 4
fix
qt
y oo −=∂∂
+∂∂
fixy
ymt
y omooo
o o −=∂∂
+∂∂ −1α
Kinematic routing methods
Governing equations Contd…
Lecture 3: Kinematic channel modeling
Module 4
Representative of collectors or stream channels
Triangular
Rectangular
Trapezoidal
Circular
These are completely characterized by slope, length, cross-sectional
dimensions, shape and Manning’s n value.
Kinematic channel modeling
Module 4
Basic channel shapes and their variations
Module 4
• The basic forms of the equations are similar to the overland flow Eq. (Eqs.4 .10 and 4.12). For stream channels or collectors,
Equations of kinematic channel modeling
………..4.13
……….4.14
where,
Ac= cross sectional flow area (ft2)
Qc= discharge
qo= overland inflow per unit length (cfs/ft)
αc, mc= kinematic wave parameter for the particular channel
Module 4
occ q
xQ
tA
=∂∂
+∂∂
cmccc AQ α=
shape αc mc
Triangular 4/3
Square 4/3
Rectangular 5/3
Trapezoidal Variable, function of A and W
Circular 5/4
Kinematic channel parameters
Module 4
3/1
2194.0
+ zz
ns
ns72.0
( )3/249.1 −Wn
s
( )6/1804.0cD
ns
• Determine αc and mc for the case of a triangular prismatic channel
11
ZZ
yc
Example Problem
Module 4
Solution
and yc = channel depth
Wetted perimeter =
hydraulic radius =
Substituting these into manning’s Eq. given by
2cc ZyAArea ==
c
c
PA
R =
Module 4
212 zyP cc +=
3/2
3/549.1
c
cc P
As
nQ =
Example Problem Contd…
From Eq.14, . Therefore,
and mc= 4/3
Module 4
( )( ) 3/123/2
3/103/5
159.149.1
Zy
yZs
nQ
c
cc
+=
( ) 3/423/1
2194.0
cc ZyZ
Zsn
Q
+
=
( ) 3/43/1
2194.0
cc AZ
Zsn
Q
+
=
cmccc AQ α=
3/1
2194.0
+
=Z
Zsncα
Example Problem Contd…
Highlights in the module
This module presents the concept of kinetic wave model which assumes the
that the weight or gravity force of flowing water is simply balanced by the
resistive forces of bed friction
The brief introduction to St. Venant equations is provided in this module,
whereas, the complete part of this is covered in module-6.
Module 4
Flood Routing
Prof. Subhankar KarmakarIIT Bombay
Module 54 Lectures
The objective of this module is to introduce the concepts and
methods of lumped and distributed flood routing along with
an insight into Muskingum method.
Module 5
Topics to be covered
Lumped flow routing
Level pool method
Kinematic wave/Channel routing
Muskingum method
Distributed Flow routing
Diffusion wave routing
Muskingum-Cunge method
Dynamic wave routing
Module 5
Lecture 1: Introduction to flood routing
Module 5
Flood Routing
“Flood routing is a technique of determining the flood
hydrograph at a section of a river by utilizing the data of
flood flow at one or more upstream sections.”
( Subramanya, 1984)Module 5
Applications of Flood Routing
Flood: Flood Forecasting Flood Protection Flood Warning
Design:Water conveyance (Spillway) systemsProtective measuresHydro-system operation
Water Dynamics:Ungauged riversPeak flow estimationRiver-aquifer interaction
For accounting changes in flow hydrograph as a flood wave passes downstream
Module 5
Types of flood routing
Lumped/hydrologic Flow f(time)
Continuity equation and Flow/Storage relationship
Distributed/hydraulic Flow f(space, time)
Continuity and Momentum equations
Module 5
Flow Routing Analysis
It is a procedure to determine the flow hydrograph at a point on a watershed from a
known hydrograph upstream.
Upstream hydrograph
Inflow)( =tI
Inflow
Q TransferFunction
Outflow)( =tQDownstream hydrograph
OutflowQ
Module 5
Flow Routing Analysis Contd…
As flood wave travels downstream, it undergoesPeak attenuation
Translation
Q
tTp
Qp
Q
tTp
Qp
Q
tTp
Qp
Module 5
Flood Routing Methods
Lumped / Hydrologic flow routing:Flow is calculated as a function of time alone at a particular location.
Hydrologic routing methods employ essentially the equation of continuity and
flow/storage relationship
Distributed / Hydraulic routing:Flow is calculated as a function of space and time throughout the system
Hydraulic methods use continuity and momentum equation along with the
equation of motion of unsteady flow (St. Venant equations).
Module 5
Hydrologic routing
1. Level pool method (Modified Puls)
Storage is nonlinear function of Q
Reservoir routing
2. Muskingum method
Storage is linear function of I and Q
Channel routing
3. Series of reservoir models
Storage is linear function of Q and its time derivatives
Module 5
Continuity equation for hydrologic routing
Flood hydrograph through a reservoir or a channel reach is a gradually varied
unsteady flow. If we consider some hydrologic system with input I(t), output Q(t), and
storage S(t), then the equation of continuity in hydrologic routing methods is the
following: Change in storage
Change in time
Module 5
Rate change of flow storage can be also represented by this following equation:
Even if the inflow hydrograph, I(t) is known, this equation cannot be solved directly
to obtain the outflow hydrograph, Q(t), because both Q and S are unknown. A
second relation, the storage function is needed to relate S, I, and Q. The particular
form of the storage equation depends on the system: a reservoir or a river reach.
Change in storage
Change in time
Contd..
Module 5
Continuity equation for hydrologic routing
Lecture 2: Level pool routing and modified Pul’s method
Module 5
Hydrologic flow routing
When a reservoir has a horizontal water surface elevation, the storage function is a
function of its water surface elevation or depth in the pool. The outflow is also a
function of the water surface elevation, or head on the outlet works.
S= f(O)
where S= storage and O= Outflow
Module 5
1. Level Pool Routing
1. Level Pool Routing Contd..
I= inflowQ= outflowS =storaget=time
Q
S
Module 5
The peak outflow occurs when the outflow hydrograph intersects the inflow hydrograph.
1. Level Pool Routing Contd…
Maximum storage occurs when
As the horizontal water surface is assumed in the reservoir, the reservoir
storage routing is known as Level Pool Routing. The outflow from a reservoir is a
function of the reservoir elevation only. The storage in the reservoir is also a
function of the reservoir elevation.
Module 5
Hydrologic flow routing
In a small time interval the difference between the total inflow and outflow in a
reach is equal to the change in storage( ) in that reach
Where = average inflow in time t, = average outflow in time t. If suffixes 1 and
2 denote the beginning and end of the time interval t then the above equation
becomes
Module 5
1. Level Pool Routing Contd…
Hydrologic flow routing
The time interval should be sufficiently short so that the inflow and out flow hydrographs can be assumed to be straight lines in that time interval.
Module 5
1. Level Pool Routing Contd…
Hydrologic flow routing
In reservoir routing, the following data are known:
(i) Elevation vs Storage
(ii) Elevation vs Outflow discharge and hence storage vs outflow discharge
(iii) Inflow hydrograph, and
(iv) Initial values of inflow, outflow O, and storage S at time t = 0.
Module 5
1. Level Pool Routing Contd…
Hydrologic flow routing
A variety of methods are available for routing of floods through a reservoir. All of them use the general equation but in various rearranged manners.
Pul’s Method: This is a semi-graphical method.
All the terms on the left hand side are known and hence right hand side at the end of
the time step . Since S = f(h) and O = f(h), the right hand side is a function of
elevation h for a chosen time interval .Graphs can be prepared for h vs O, h vs S
and h vs . .
Module 5
Modified Pul’s Method
1. Level Pool Routing Contd…
Hydrologic flow routing
STEPSFor practical use, this semi-graphical is very convenient:1. From the known storage-elevation and discharge-elevation data, prepare a
curve of vs elevation. Here is any chosen interval, approximately
20 to 40% of the time rise of the inflow hydrograph.
2. On the same plot prepare a curve of outflow discharge vs elevation.
3. The storage, elevation and outflow discharge at the starting of routing are
known. For the first time interval , , and are known and
hence the term is determined form the Pul’s method.
4. The water –surface elevation corresponding to is found by using the
plot of step (1). The outflow discharge Q2 at the end of the time step is
found from the plot of step (2).
Module 5
Level Pool Routing
Modified Pul’s Method Contd…
STEPS
4. The water –surface elevation corresponding to is found by using
the plot of step (1). The outflow discharge Q2 at the end of the time step
is found from the plot of step (2).
5. Deducting from gives for the beginning of
the next step.
6. The procedure is repeated till the entire inflow hydrograph is routed.
Module 5
Modified Pul’s Method Contd…
Level Pool Routing
Lecture 3: Channel routing methods
Module 5
2. Channel Routing
In very long channels the entire flood wave also travels a considerable distance
resulting in a time redistribution and time of translation as well. Thus, in a
river, the redistribution due to storage effects modifies the shape, while the
translation changes its position in time. In reservoir, the storage is a unique
function of the outflow discharge S = f(O).
Storage in the channel is a function of both outflow and inflow discharges and
hence a different routing method is needed. The water surface in a channel
reach is not only parallel to the channel bottom but also varies with time.
Module 5
Hydrologic flow routing
The total volume in storage for a channel reach having a flood wave can be
considered as prism storage + wedge storage.
Prism storage: The volume that would exist if uniform flow occurred at the
downstream depth i.e. the volume formed by an imaginary plane parallel to the
channel bottom drawn at the outflow section water surface.
Wedge storage: It is the wedge like volume formed between the actual water
surface profile and the top surface of the prism storage. At a fixed depth at a
downstream section of a river reach the prism storage is constant while the wedge
storage changes from a positive value at an advancing flood to a negative value
during a receding flood.
Module 5
2. Channel Routing Contd…
Hydrologic flow routing
Prism Storage: It is the volume
that would exits if uniform flow
occurred at the downstream
depth, i.e. the volume formed by
an imaginary plane parallel to the
channel bottom drawn at the
outflow section water surface.
Module 5
2. Channel Routing Contd…
Hydrologic flow routing
Wedge storage : It is the wedge-like volume formed between the actual water surface profile and the top surface of the prism storage.
Module 5
2. Channel Routing Contd…
Hydrologic flow routing
At a fixed depth at a downstream section of a river reach, the prism storage is constant
while , the wedge storage changes from a positive value for advancing flood to a
negative value during a receding flood.
Total storage in the channel reach can be expressed as :
where k and x are coefficients and m= a constant exponent . It has been found that
m varies from 0.6 for rectangular channels to a value of about 1.0 for natural
channels, Q = outflow
Module 5
Hydrologic flow routing
2. Channel Routing Contd…
Assuming that the cross sectional area of the flood flow section is directly
proportional to the discharge at the section, the volume of prism storage is equal
to KQ where K is a proportionality coefficient, and the volume of the wedge
storage is equal to KX(I- Q), where X is a weighing factor having the range 0 < X
< 0.5. The total storage is therefore the sum of two components
)( QIKXKQS −+=
It is known as Muskingum storage equation representing a linear model for routing flow in streams.
Module 5
Muskingum Method
Channel routing
QI −
IQ
II
IQ −
I Q
AdvancingFloodWaveI > Q
RecedingFloodWaveQ > I
KQS =Prism
)(Wedge QIKXS −=
K is a proportionality coefficient,
X is a weighing factor on inflow versus
outflow (0 ≤ X ≤ 0.5)
X = 0.0 - 0.3 Natural stream
)( QIKXKQS −+=
])1([ QXXIKS −+=
Module 5
Muskingum Method Contd…
Channel routing
])1([ QXXIKS −+=
The value of X depends on the shape of the modeled wedge storage. It is zero for
reservoir type storage (zero wedge storage or level pool case S = KQ) and 0.5
for a full wedge. In natural streams mean value of X is near 0.2. The parameter K
is the time of travel of the flood wave through the channel reaches also
known as storage time constant and has the dimensions of time.
Module 5
Muskingum Method Contd…
Channel routing
From the Muskingum storage equation, the values of storage at time j and j+1 can be written as
and
So, change in storage over time interval ∆t is,
From the continuity equation the storage for the same time interval ∆t is,
Module 5
Muskingum Method Contd…
Channel routing
Equating these two equations,
Collecting similar terms and simplifying
This is the Muskingum’s routing equation for channels Module 5
Muskingum Method Contd…
Channel routing
Muskingum’s routing equation for channels:
where
For best results, the routing interval ∆t should be so chosen that K>∆t>2KX. If ∆t<2KX, the coefficient C1 will be negative. Generally negative values of coefficients are avoided by choosing appropriate values of ∆t.
Module 5
Muskingum Method Contd…
Channel routing
To use Muskingum equation to route a given inflow hydrograph through a
channel reach:
K , X and Oj should be known.
Procedure:
(i)knowing K and X, select an appropriate value of t
(ii) calculate C1, C2, and C3
(iii) starting from the initial conditions known inflow, outflow calculate the outflow for
the next time step.
(iv) Repeat the calculations for the entire inflow hydrograph.
Module 5
Muskingum Method Contd…
Channel routing
Lecture 4: Hydraulic routing
Module 5
Hydraulic/Distributed flow routing
Flow is calculated as a function of space and time throughout the system
Hydraulic methods use continuity and momentum equation along with the
equation of motion of unsteady flow (St. Venant equations).
St. Venant Equations (Refer to Module 6 for more details)
Kinematic wave routing
Diffusion wave routing
Muskingum-Cunge method
Dynamic wave routing
Module 5
It is the relationship between the Muskingum method and the Saint-Venant equations.
Inflow-Outflow Equation:
The constants C0, C1 and C2 are functions of wave celerity, c.
Qdischarge and y depth of flow
Muskingum-Cunge method
Diffusion wave routing
Module 5
,
dydA
dydQ
dAdQc ==
tOIIO 210 CCC ttttt ++= ∆+∆+
where,
Q0 = Reference discharge,
S0 = Reach Slope,
QB = Baseflow
Qp = Peak flow taken from the inflow hydrograph
Module 5
Muskingum-Cunge method Contd…
Diffusion wave routing
∆
−=xSTc
QX***
121
0
0
( )BpB QQQQ −+= 50.00
Dynamic Wave Routing
Flow in natural channels is unsteady, non-uniform with junctions, tributaries,variable cross-sections, variable resistances, variable depths, etc. The completeSt.Venant equation represents the dynamic wave routing. (Refer to Module 6 for moredetails)
Valley storage
Prismstorage
Wedgestorage
Non-conservative form of continuity equation
Module 5
∂∂
+
∂∂
+∂∂
=xyV
xVy
ty0
Dynamic Wave Routing Contd…
Momentum equation considering all relevant forces acting on the system:
Local acceleration
term
Convective acceleration
term
Pressure force term
Friction force term
Module 5
0)(11 2
=−−∂∂
+
∂∂
+∂∂
fo SSgxyg
AQ
xAtQ
A
Gravity force term
Example Problem
Given:
Inflow hydrograph
K = 2.3 hr, X = 0.15, ∆t = 1 hour, Initial Q = 90 cfs
Find:
Outflow hydrograph using Muskingum routing method
Module 5
5927.01)15.01(3.2*21)15.01(*3.2*2
)1(2)1(2
3442.01)15.01(3.2*2
15.0*3.2*21)1(2
2
0631.01)15.01(3.2*2
15.0*3.2*21)1(2
2
3
2
1
=+−−−
=∆+−∆−−
=
=+−
+=
∆+−+∆
=
=+−
−=
∆+−−∆
=
tXKtXKC
tXKKXtC
tXKKXtC
Period Inflow (hr) (cfs)
1 93 2 137 3 208 4 320 5 442 6 546 7 630 8 678 9 691
10 675 11 634 12 571 13 477 14 390 15 329 16 247 17 184 18 134 19 108 20 90
Example Problem Contd…
jjjj QCICICQ 32111 ++= ++
C1 = 0.0631, C2 = 0.3442, C3 = 0.5927
Period Inflow C1Ij+1 C2Ij C3Qj Outflow(hr) (cfs) (cfs)
1 93 0 0 0 90
2 137 9 32 53.343 94.343
3 208 13 47 55.9171 115.9171
4 320 20 72 68.70406 160.7041
5 442 28 110 95.2493 233.2493
6 546 34 152 138.2469 324.2469
7 630 40 188 192.1811 420.1811
8 678 43 217 249.0413 509.0413
9 691 44 233 301.7088 578.7088
10 675 43 238 343.0007 624.0007
11 634 40 232 369.8452 641.8452
12 571 36 218 380.4217 634.4217
13 477 30 197 376.0217 603.0217
14 390 25 164 357.411 546.411
15 329 21 134 323.8578 478.8578
16 247 16 113 283.819 412.819
17 184 12 85 244.6778 341.6778
18 134 8 63 202.5124 273.5124
19 108 7 46 162.1108 215.1108
20 90 6 37 127.4962 170.4962
0
100
200
300
400
500
600
700
800
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Dis
char
ge (c
fs)
Time (hr)
Inflow Outflow
Exercise Problem
An inflow hydrograph is measured for a cross section of a stream. Compute the
outflow hydrograph at a point five miles downstream using the Muskinghum
method . Assuming K = 12hr, x=0.15, and outflow equals inflow initially. Plot the
inflow and outflow hydrograph.
Time Inflow (cfs)9:00A.M. 503:00P.M 75
9:00 P.M. 1503:00A.M. 4509:00A.M. 10003:00P.M. 8409:00P.M. 7503:00A.M. 6009:00A.M. 3003:00P.M. 1009:00P.M. 50
Module 5
Highlights in the Module
Flood routing is a technique of determining the flood hydrograph at a section
of a river by utilizing the data of flood flow at one or more upstream sections
As a flood wave travels downstream, it undergoes:
Peak attenuation
Translation Types of flood routing
Lumped/hydrologic
Distributed/hydraulic
Lumped / Hydrologic flow routing
Flow is calculated as a function of time alone at a particular location. Equation of continuity and flow/storage relationship
Module 5
Highlights in the Module Contd…
Hydrologic flow routing methods
Level pool method (Modified Puls)
Channel routing\Muskingum method
Series of reservoir models
Distributed / Hydraulic routing
Flow is calculated as a function of space and time throughout the system
Hydraulic methods use continuity and momentum equation along with the
equation of motion of unsteady flow (St. Venant equations).
Module 5
Distributed / Hydraulic routing methods
Diffusion wave routing
Muskingum Cunge method
Dynamic wave routing
Complete solution to St.Venant equations
Module 5
Highlights in the Module Contd…
Prof. Subhankar KarmakarIIT Bombay
Equations Governing Hydrologic and Hydraulic Routing
Module 65 Lectures
Objectives of this module is to understand the physical phenomena behind the Reynolds transport theorem and
Saint Venant equations.
Module 6
Topics to be covered
Reynolds Transport Theorem
Control Volume Concept
Open Channel Flow
Saint Venant Equations
Continuity Equation
Momentum Equation
Energy Equation
Module 6
Lecture 1: Reynolds transport theorem and open channel flow
Module 6
Fluids Problems-Approaches
1. Experimental Analysis
2. Differential Analysis
3. Control Volume Analysis
Looks at specific regions, rather than specific masses
Module 6
Reynolds' transport theorem (Leibniz-Reynolds' transport theorem) is a 3-D generalization of the Leibniz integral rule. The theorem is named after Osborne Reynolds (1842–1912).
Control volume: A definite volume specified in space. Matter in a control volume can change with time as matter enters and leaves its control surface.
Extensive properties (B) : Properties depend on the mass contained in a fluid, Intensive properties (β) : Properties do not depend on the mass.
Osborne Reynolds
Reynolds Transport Theorem
dB or dB dmdm
β β= =
mass (m) momentum
P.E. (mgh) K.E. (1/2 mv2)
1)(==
dmmdβ )( vm
gh=β
Module 6
vdm
vmd==
)(β2
21 v=β
This theorem applies to any transportable property,
including mass, momentum and energy.
Module 6
Reynolds Transport Theorem:
The total rate of change of any extensive property B (=βdm = βρd∀) of a system
occupying a control volume C.V. at time ‘t’ is equal to the sum of:
a) the temporal rate of change of B within the C.V.
b) the net flux of B through the control surface C.S. that surrounds the C.V.
∫∫∫∫∫ +∀∂∂
=....
.sc
relvc
dAVdtdt
dB βρβρ
Reynolds Transport Theorem Contd…
Reynolds Transport Theorem can be applied to a control volume of finite size
No flow details within the control volume is required
Flow details at the control surfaces is required
We will use Reynolds Transport Theorem to solve
many practical fluids problems
Here, control volume is the sum of I & II
fluid particles at time ‘t’o fluid particles at time ‘t+∆t’
Module 6
Reynolds Transport Theorem Contd…
III III
- Control volume position occupied by fluid at time ‘t’
- Position occupied by fluid at time ‘t+∆t’
- Position occupied by fluid at time ‘t’, but not at ‘t+∆t’- Position occupied by fluid at both the time ‘t’ and ‘t+∆t’
- Position occupied by fluid at time ‘t+∆t’, but not at ‘t’
III
III
Fixed frame in space(upper and lower boundaries are impervious)
Finally, control volume at time ‘t’ I&II and at time ‘t+∆t’ II&III
Let us consider the following system,
Module 6
1st term 2nd term1st term is equivalent to the change in extensive property stored in control volume (as ∆t 0)
Module 6
( ) ( )[ ]
( )
( ) ( )[ ] ( ) ( )[ ]{ }tIttIIItIIttIItsys
tIIIttIIIIItsys
BBBBtdt
dB
eqFrom
dincontainedfluidtheofmassdm
volumeelementaldddmwhere
ddmdBHere
BBBBtdt
dB
Lt
Lt
−−−∆
=
∀=
=∀∀
=
∀==
+−+∆
=
∆+∆+→∆
∆+→∆
1
)1.6.(
,
)1.6...(1
0
0
ρ
ρββ
Extensive property in the control volume,
Here, area vector is always normally outward to the surface
dA
control surface
V
θ∆l
Module 6
( )
dAl
ldA
ttimeindAthroughgpasfluidofVolume
vectorvelocityv
dAmagnitudeofvectorAreaAd
.cos
cos
""""sin
θ
θ
∆−=
∆−=
∆
=
=
Module 6
( )
( )
( )
)(
)(
:
cos
sin
..
surfacecontrolthroughpropertyextensiveofoutflowNet
volumecontroltheinpropertiesextensiveofchangeofRate
fluidofpropertyextensiveofchangeofrateTotaldtdB
changeShapepropertiesextensiveinChangeBdtd
onlypropertiesextensiveinChangeBt
Note
ldA
surfacecontrolthethroughgpasfluidofvolumeTotal
sys
sc
+=
=⇒
+⇒
⇒∂∂
∆−= ∫∫ θ
:sAssumptiontconsisi tan)( ρ
Module 6
( )
)(inf)()()(,
),()(.
0.
0,
.var
)2.6......(0.
0
,,.
....
....
..
....
velowtIandveoutflowtQstorageS
wheretItQdAVandSd
dAVddtd
ddtdflowstatesteadyFor
flowunsteadydensityiableofequationtheisThis
dAVddtd
dtmassd
dtdB
massofonconservatioflawperasNowmassbetoBConsider
scvc
scvc
vc
scvc
sys
−=+==
−==∀∴
=+∀
=∀−
=+∀∴
==
∫∫∫∫∫
∫∫∫∫∫
∫∫∫
∫∫∫∫∫
ρ
ρρ
Reynolds transport equation becomes,
Unsteady State
Steady State
Module 6
.
,int
,1
,)(
)()(
)()(
)()(,0
.,0)()(
0)()(.
1
1
1
1
)1()1(1
jjj
jjj
j
j
jjjj
tj
tj
tj
tj
S
S
tandttimethebetweensystemthefromOutflowQ
tandttimethebetweensystemtheoInflowI
jtimeatstorageS
jtimeatstorageSwhereQISS
dttQdttIdS
dttQdttIdS
tQtIthendtdS
provedhencetQtIdtdSor
tItQdtdSei
j
j
−
−
−
−
∆
∆−
∆
∆−
=
=
−=
=
−=−
−=∴
−=∴
==∴
=−=
=−+
∫∫∫−
flowsteadyandtconsisii tan)( ρ
∑∫ ⋅+∀=CSCV
sys ddtd
dtdB
AV
βρβρ
∑ ⋅=CS
sys
dtdB
AV
βρ
Module 6
,
)()(
),4.6(
)5.6......(),3.(
)4.6......()(,2
)3.6......()(
,1
221102
221101
1
1101
2212
1101
Thus
QIQISSorQIIQSS
inSofvaluengSubstituti
IQSSEqFrom
QISSj
QISS
jIf
−+−+=−=−+−
+−=
−=−=
−=−
=
∑=
−+=j
tttj QISS
10 )(
An open channel is a waterway, canal or conduit in which a liquid flows
with a free surface.
In most applications, the liquid is water and the air above the flow is
usually at rest and at standard atmospheric pressure.
udel.edu/~inamdar/EGTE215/Open_channel.pdf
Open channel flow
Module 6
Pipe flow Open channel flow
Flow driven by Pressure work Gravity(i.e. Potential Energy)
Flow cross-section Known (Fixed by
geometry)
Varies based on the depth of flow
Characteristic flow
parameters
Velocity deduced from
continuity equation
Flow depth and velocity deduced by
solving simultaneously the
continuity and momentum
equations
Specific boundary
conditions
Atmospheric pressure at the water
surface
Pipe flow vs. Open channel flow
(Source:www.uq.edu.au/~e2hchans/reprints/b32_chap01.pdf)
Different flow conditions in an open channel
Section 1 – rapidly varying flowSection 2 – gradually varying flowSection 3 – hydraulic jump
Section 4 – weir and waterfallSection 5 – gradually varyingSection 6 – hydraulic drop due to change in channel slope
Module 6
Open Channel Flow
Unsteady Steady
Varied Uniform Varied
Gradually
Rapidly
Gradually
Rapidly
(NPTEL, Computational Hydraulics)
Module 6
Different Types of flow in an open channel
1. Steady Uniform flow
2. Steady Gradually-varied flow
3. Steady Rapidly-varied flow
4. Unsteady flow
Module 6
Case (1) – Steady uniform flow:
Steady flow is where there is no change with time, ∂/∂t = 0.
Distant from control structures, gravity and friction are in balance, and if the
cross-section is constant, the flow is uniform, ∂/∂x = 0
Case (2) – Steady gradually-varied flow:
Gravity and friction are in balance here too, but when a control is introduced
which imposes a water level at a certain point, the height of the surface varies
along the channel for some distance.
i.e. ∂/∂t = 0, ∂/∂x ‡ 0.
Case (3) – Steady rapidly-varied flow:
Here depth change is rapid.
Case (4) – Unsteady flow:
Here conditions vary with time and position as a wave traverses the waterway.
Module 6
Types of Open Channel Flows
Module 6
Hydraulic Routing
Hydraulic routing
Momentum Equation
Physics of watermovement
Hydraulic routing is intended to describe the dynamics of the water or flood wave
movement more accurately
Hydrological routing
Continuity equation + f (storage, outflow, and possibly inflow) relationships
assumed, empirical, or analytical in nature. Eg: stage-discharge relationship.
Module 6
Dynamic Routing- Advantages
Higher degree of accuracy when modeling flood situations because it
includes parameters that other methods neglect.
Relies less on previous flood data and more on the physical properties of
the storm. This is extremely important when record rainfalls occur or other
extreme events.
Provides more hydraulic information about the event, which can be used to
determine the transportation of sediment along the waterway.
Module 6
Lecture 2: Navier-Stokes and Saint Venantequations
Module 6
Navier-Stokes Equations
St.Venant equations are derived from Navier-Stokes Equations for shallow
water flow conditions.
The Navier-Stokes Equations are a general model which can be used to model
water flows in many applications.
A general flood wave for 1-D situation can be described by the Saint-Venant
equations.
Claude-Louis Navier Sir George Gabriel Stokes
Module 6
Navier-Stokes Equations Contd…
It consists of 4 nonlinear PDE of mixed hyperbolic-parabolic type describing
the fluid hydrodynamics in 3D.
Expression of F=ma for a fluid in a differential volume
The acceleration vector contains local and convective acceleration terms
where i: x, y, zui: u, v, wuj: u, v, w
Module 6
( )
( )
( )8.6
7.6
6.6
zww
ywv
xwu
twa
zvw
yvv
xvu
tva
zuw
yuv
xuu
tua
z
y
x
∂∂
+∂∂
+∂∂
+∂∂
=
∂∂
+∂∂
+∂∂
+∂∂
=
∂∂
+∂∂
+∂∂
+∂∂
=
j
ij
ii x
uutua
∂∂
+∂∂
=
The force vector is broken into a surface force and a body force per unit volume.
The body force vector is due only to gravity while the pressure forces and the
viscous shear stresses make up the surface forces(i.e. per unit mass).
Module 6
)11.6(1
)10.6(1
)9.6(1
∂∂
+∂∂
+∂∂
+∂∂
−+=
∂∂
+∂∂
+∂∂
+∂∂
−+=
∂∂
+∂∂
+∂∂
+∂∂
−+=
zyxzpgf
zyxypgf
zyxxpgf
zzyzxzzz
zyyyxyyy
zxyxxxxx
τττρ
τττρ
τττρ
Navier-Stokes Equations Contd…
The stresses are related to fluid element displacements by invoking the
Stokes viscosity law for an incompressible fluid.
( )
( )
( )15.6
14.6
)13.6(
12.62,2,2
∂∂
+∂∂
==
∂∂
+∂∂
==
∂∂
+∂∂
==
∂∂
=∂∂
=∂∂
=
yw
zv
zu
xw
xv
yu
xw
xv
xu
zyyz
zxxz
yxxy
zzyyxx
µττ
µττ
µττ
µτµτµτ
Module 6
Navier-Stokes Equations Contd…
Substituting eqs. 6.12-6.15 into eqs. 6.9-6.11, we get,
Module 6
notationEinsteinxx
uxpgf
zw
yw
xw
zpgf
zv
yv
xv
ypgf
zu
yu
xu
xpgf
jj
i
i
ii
zz
yy
xx
∂∂∂
+∂∂
−=
∂∂
+∂∂
+∂∂
+∂∂
−=
∂∂
+∂∂
+∂∂
+∂∂
−=
∂∂
+∂∂
+∂∂
+∂∂
−=
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
1
)18.6(1
)17.6(1
)16.6(1
νρ
νρ
νρ
νρ
Navier-Stokes Equations Contd…
The equation of continuity for an incompressible fluid
The three N-S momentum equations can be written in compact form as
Module 6
)19.6(1 2
i
jj
i
ij
ij
i gxx
uxp
xuu
tu
+∂∂
∂+
∂∂−
=∂∂
+∂∂ ν
ρ
)20.6(0
0
=∂∂
=∂∂
+∂∂
+∂∂
i
i
xu
zw
yv
xu
Navier-Stokes Equations Contd…
The Saint Venant Equations were formulated in the 19th
century by two mathematicians, de Saint Venant and
Bousinnesque.
The solution of the St. Venant equations is known as dynamic
routing, which is generally the standard to which other methods
are measured or compared. Jean Claude Saint-Venant
Joseph Valentin Boussinesq
Continuity equation:
Momentum equation:
Q-Discharge through the channelA-Area of cross-section of flowy- Depth of flowS0-Channel bottom slopeSf- Friction slope
0=∂∂
+∂∂
tA
xQ
0)(11 2
=−−∂∂
+
∂∂
+∂∂
fo SSgxyg
AQ
xAtQ
A
Saint Venant Equations
Assumptions of St. Venant Equations
• Flow is one-dimensional
• Hydrostatic pressure prevails and vertical accelerations are negligible
• Streamline curvature is small.
• Bottom slope of the channel is small.
• Manning’s and Chezy’s equation are used to describe resistance effects
• The fluid is incompressible
• Channel boundaries are considered fixed and therefore not susceptible to
erosion or deposition.
1D gradually varied unsteady flow in an open channel is given by St. Venant
equations:
Continuity Equation ( based on Conservation of Mass)
Momentum Equation ( based on Conservation of Momentum)
Module 6
In the diagrams given,Q = inflow to the control volume
q = lateral inflow
= Rate of change of flow
with distance
= Outflow from the C.V.
= Change in mass
1-D Open channel flow
dxxQQ∂∂
+
tAdx∂
∂ )(ρ
xQ∂∂
Plan View
Elevation View
Module 6
St. Venant equations
Conservation of MassIn any control volume consisting of the fluid (water) under consideration, the net
change of mass in the control volume due to inflow and outflow is equal to the net
rate of change of mass in the control volume
Continuity equation:
0=∂∂
+∂∂
tA
xQ Q-Discharge through the channel
A-Area of cross-section of flow
Module 6
Q = AV = volume water discharge [L3/T]ρQ = Mass water discharge = ρAV [M/T]
∂/∂t(Mass in control volume) = Net mass inflow rate (assuming q=0)
Continuity Equation-Derivation
Module 6
( ) ( )
( ) ( )
0
sec
arg,;0
0.
=
∂∂
+∂∂
⇒
−
==
∂∂
+∂∂
∆⇒
=∆∂
∂+∆
∂∂
∆∂
∂−=−=∆
∂∂
∆+
xQ
tA
tioncrossthethrough
edischQAVHerex
AVtAx
xx
AVxtAei
xx
AVAVAVxtA
xxx
ρ
ρρ
ρρρρ
In 1-D open channel flow continuity equation becomes,
0)(=
∂∂
+∂
∂ty
xVy
Non-conservation form (velocity is dependent variable)
0=−∂∂
+∂∂ q
tA
xQ
Conservation form
Module 6
0=∂∂
+∂∂
+∂∂
ty
xVy
xyV
Example Problem
Calculate the inlet velocity Vin from the diagram shown.
Module 6
)0025.0(1*2)0025.0(101.0*1.0
)(
0
2 gVx
AVAVdtdhA
AVAVhAdtd
ddtd
in
outoutinintank
outoutinintank
CSCV
+−=
+−=
+−=
⋅+∀=
−
∑∫
ρρρ
ρρ AV
smVin /47.4=
( )ss mvF ∆=∑
Momentum
In mechanics, as per Newton’s 2nd Law: Net force = time rate of change of momentum
Sum of forces in the s direction
Change in momentum in the s direction
mass
Velocity in the s direction
Momentum Equation
The change in momentum of a body of water in a flowing channel is equal to the resultant of all the external forces acting on that body.
Sum of forceson the C.V.
Momentum storedwithin the C.V
Momentum flowacross the C. S.
Module 6
∫∫∫∫∫∑ +∀=....
.scvc
dAVVdVdtdF ρρ
∫∫∫∫∫∑ +∀=....
.scvc
dAVVdVdtdF ρρ
This law states that the rate of change of momentum in the control volume is equal to the net forces acting on the control volumeSince the water under consideration is moving, it is acted upon by external forces which will lead to the Newton’s second law
Sum of forces on the C.V.
Momentum stored within the C.V
Momentum flow across the C. S.
Module 6
0)(11 2
=−−∂∂
+
∂∂
+∂∂
fo SSgxyg
AQ
xAtQ
A
Conservation of Momentum
Applications of different forms of momentum equation
Kinematic wave: when gravity forces and friction forces balance each other
(steep slope channels with no back water effects)
Diffusion wave: when pressure forces are important in addition to gravity and
frictional forces
Dynamic wave: when both inertial and pressure forces are important and
backwater effects are not negligible (mild slope channels with downstream
control)
Module 6
The three most common approximations or simplifications are:KinematicDiffusionQuasi-steady models
Approximations to the full dynamic equations
Kinematic wave routing:
Assumes that the motion of the hydrograph along the channel is controlled by
gravity and friction forces. Therefore, uniform flow is assumed to take place. Then
momentum equation becomes a wave equation:
where Q is the discharge, t the time, x the distance along the channel, and c the
celerity of the wave (speed).
A kinematic wave travels downstream with speed c without experiencing any
attenuation or change in shape. Therefore, diffusion is absent.
0=∂∂
+∂∂
xQc
tQ
The diffusion wave approximation includes the pressure differential term
but still considers the inertial terms negligible; this constitutes an
improvement over the kinematic wave approximation.
The pressure differential term allows for diffusion (attenuation) of the
flood wave and the inclusion of a downstream boundary condition which
can account for backwater effects.
This is appropriate for most natural, slow-rising flood waves but may lead
to problems for flash flood or dam break waves
xySS f ∂∂
−= 0
Module 6
Diffusion wave routing
It incorporates the convective acceleration term but not the local
acceleration term, as indicated below:
In channel routing calculations, the convective acceleration term and
local acceleration term are opposite in sign and thus tend to negate each
other. If only one term is used, an error results which is greater in
magnitude than the error created if both terms were excluded
(Brunner, 1992).
Therefore, the quasi-steady approximation is not used in channel routing.
Module 6
Quasi-Steady Dynamic Wave Routing
)()(0 xgVV
xySS f ∂
∂−
∂∂
−=
Lecture 3: Derivation of momentum equation
Module 6
(a) (b)
Diagrammatic representation of (a) Isovels and (b) Velocity profile in an open channel flow
60
Isovels
10080
40
Velocity Distribution in an Open Channel Flow
60
Module 6
γh
h
γ(h-anh/g)
Gradually Varied Flow
Convex surface
an : acceleration component acting normal to the streamlines
For horizontal surface
For curved surface
γanh/g)
Convex Curvi-linear flowHydrostatic pressure distribution in a curved surface is,
yan
h
Module 6
γh
h
Concave surface
For horizontal surface
For curved surface
γanh/g)
h
Concave Curvi-linear flow
an
y
Gradually Varied Flow (Contd..)
Module 6
I, Q
y1 y2
t
Note: y=y(x,t), Depth of flow varies with distance and timeA = A (x,y) , Area of flow
Information can be known about:1) Critical inflow hydrograph 2) How much area may be
flooded for critical cases
For mass (m) 1)(==
dmmdβ
From Reynold’s transport theorem, we have
Module 6
Gradually Varied Flow (Contd..)
Module 6
( )
,var
0
,,
.
flowunsteadydensityiableofequationtheisThis
dtmassd
dtdB
massofonconservatioflawperasNow
massbetoBConsider
sys ==
tconsisand tanρ
txoffunctionareiablestheallthatindicatest
dAVdt scvc
&var
0.....
∂∂
=+∀∂∂
⇒ ∫∫∫∫∫ ρρ
0.....
=+∀∴ ∫∫∫∫∫scvc
dAVddtd ρρ
Control volume (A.dx)
-Longitudinal Section of the Channel
Module 6
( ) )21.6(....0 ∫∫∫∫ ++∂
∂=
outletinlet
dAVdAVt
dxAor ρρρ
yηdη
1 2dx
q/2
Component in x-direction
q/2
-Top view of the open channel section
-Cross-sectional view of a compound channel
Total contribution is q.dx
Module 6
xQ
tAq
∂∂
+∂∂
=
Basic equation or Conservative form of
continuity equation (Applicable for kinematic &
non-prismatic channels)
Module 6
( ) ( )
( )
( )
orxQq
tA
dxbysidesbothDividing
dxxQqdx
tdxA
dxxQqdx
tdxA
dxxQQqdxQ
tdxA
aswrittenrebecanEquation
∂∂
−−∂∂
=
∂∂
−−∂
∂=
∂∂
+−∂
∂=
∂∂
+++−∂
∂=
−
0
,''
,..
..
..0
,)21.6(
ρ
ρρ
ρρρ
ρρρ
dη
η
dy
y
y- η
B’
Module 6
( )
( ) ( )
0,
,;,
.,
=∂∂
∂∂
∂∂
+
∂∂
+
∂∂
∂∂
=
==∂∂
+∂∂
+∂∂
=
∂∂
+∂∂
=
xAchannelprismaticFor
xy
yAV
xVA
ty
yAqor
txyytyfAHerexAV
xVA
tAqor
xVA
tAqNow
Module 6
( ) ( )
∂∂
+
∂∂
+∂∂
=
==
∂∂
+
∂∂
+
∂∂
=
==
=⇒=
∂∂
++
∂∂
+
∂∂
=
≅∂∂=
xyV
xVD
tyor
channeltheofwidthunitBqgAssuxyBV
xVDB
tyBqor
txyytyfAHereDBABADdepthhydraulicNow
channelprismaticforxyBV
xVA
tyBqor
ByASo
dyBdA
0
)(1'&0,min
'.'.'
,;,'.'/)(,
,'0'
',
'
Valley storage
Prismstorage
Wedgestorage
Non-conservative form of continuity equation
Now, from Reynolds transport theorem:
From conservation law of momentum(Newton’s 2nd law of motion) ∑= F
dtdB
momentum )( vm
Module 6
or
yByBBADei
lengthchannelthethroughoutuniformsameisdepthIf
=== '/'.'.,.
,)(
∂∂
+
∂∂
+∂∂
=xyV
xVy
ty0
∫∫∫∫∫ +∀∂∂
=....
.scvc
dAVdtdt
dB βρβρ
vdm
vmd==
)(β
Unsteady Non-uniform flow
Steady Non-uniform flow
0∑ =FFor steady uniform flow,
Module 6
∫∫∫∫∫∑ +∀∂∂
=∴....
.scvc
dAVVdVt
F ρρ
Forces acting on the C.V.
-Plan View
-Elevation View
Module 6
Fg = Gravity force due to weight of water in the C.V.
Ff = friction force due to shear stress along the bottom and sides of the C.V.
Fe = contraction/expansion force due to abrupt changes in the channel cross-
section
Fw = wind shear force due to frictional resistance of wind at the water surface
Fp = unbalanced pressure forces due to hydrostatic forces on the left and
right hand side of the C.V. and pressure force exerted by banks
Forces acting on the C.V. (Contd..)
Module 6
γ.η γ.(y-η)
a) Hydrostatic Force
dη
η
dy
yy- η
B’
b
Area of the Elemental strip=b.d η
Force=area*hydrostatic pressure
or dF=(b.d η)*γ.(y-η)
Module 6
( ) ( )y
F b.d y dη
η γ η η=
= −∫10
Module 6
( )
( ) ( )
p
y
p
F&F F dxx
&F F F
F FF (F dx ) dxx x
F b.d y dxx
and y y( x,t ) (u sin g Leibnitz rule)
η
η γ η=
∂ = + ∂ = −
∂ ∂ = − + =− ∂ ∂ ∂ ⇒ = − − ∂
=
∫
12 1
1 2
1 11 1
0
1 2
1F dxxFF
∂∂
+ 11
a) Hydrostatic Force Contd…
( ) ( )
{ } ( )
b( t )
a( t )
b( t )
a( t )
[General rule :
F t x,t dx
d dF(t ) x,t dxdt dt
φ
φ
=
=
∫
∫
Module 6
( ) ( ) ( )
( ) ( ) ( ) ( ) ( )
( ) ( )
b( t )
a( t )
y
p
y
d x,t db( t ) da( t )dx b( t ),t . a( t ),t .dt dt dt
Therefore,d d
F b y d b y y b yx dx dx
(b)b y y d dxx x
yb
η
η
η
η
φφ φ
ηγ η η γ γ
γ η η η
γ
=
=
=
=
= + −
∂ = − − − − + − ∂ ∂ ∂ = − − + − ∂ ∂
∂= −
∂
∫
∫
∫
0
0
00
( )
( )
y
y y
by d dxx x x
y bb d y d dxx x
η γ η η
γ η γ η η
∂ ∂ − + − ∂ ∂ ∂ ∂ = + − ∂ ∂
∫
∫ ∫
0
0 0
( )
( ) dxdxbydxA
xy
dxdxbydxdb
xy
y
yy
∂∂
−−
∂∂
−=
∂∂
−−
∂∂
−=
∫
∫∫
ηηγγ
ηηγηγ
0
00
b) Boundary Reaction
Module 6
( )
.
..
,
)2()2(sec
)1()1(sec.
alsoondistributipressurechydrostatitheinChange
dxdxbdx
xdbdx
xA
areaelementalinChange
tionatareaElementalAdxx
A
tionatAdbareaElementary
element
elementelement
element
∴∂∂
=
∂∂
=
∂∂
=
∴
−=+
∂∂
−==
ηη
η
γy
y
y-η
η
γ(y-η)
dη
dη
b1
q/2
q/2
b
dx
Module 6
( )
( )
y
y
(bd ) y
In similar way,
(bd )dx. . y ,x
additional hydrostatic forcefor change in width of the channel.
η γ η
η γ η
−
∂−
∂
∫
∫
0
0
b) Boundary Reaction Contd…
W
Module 6
c) Body Force
θ
w.sinθ
θ
γ
sin.
..,
,tan)(
WFdxAWWeight
tconsbeAareaLet
g =∴
=∴
Module 6
( )
depthmeanhydraulicorradiushydraulicRwhere
SRSPAor
balancedentirelyareforcestwoThesesAssumption
stressshearperimeterwettedPwhereSdxAdxPissidesandbottomchannelforforceShear
SdxAForSandofvaluessmallveryFor
g
=
=
=
===
=
=≅
,....
:
,.....
,
...
tantansin,
000
000
0
0
γγτ
τγτ
γ
θθθθ
c) Body Force Contd…
.,, 0
hereflowuniformunsteadyconsidertoneedweHoweverSSflowuniformsteadyFor f=
d) Frictional Force
e) Wind Force
Module 6
( )
ff
f
f
SdxAFor
SRdxPdxPforceShear
SR
...
......
..
0
0
γ
γτ
γτ
−=
−=
−=∴
=
frrf
w
ww
w
WVVC
channeltheoflengththethroughoutuniformisitandBiswidthtopIfdxBwindforceF
bestressshearwindLet
ρρ
τ
ττ
−=−=
==
2...
)(..
.,
.,
veissoveisVveisVIf
veissoveisVveisVIf
wrr
wrr
+−⇒−
−+⇒+
τ
τ
ffw
f
f
WdxBWdxBFfactorstressshearwindW
tcoefficienstressshearwindC
...)(. ρρ −=−=∴
=
=
f) Eddy Losses:
( )2
2
2
2
22
2
2
)(
∂∂
=∂∂
=
∂∂
=
∂∂
∝
⇒
AQ
xgKv
xgKS
gv
xKSor
gv
xS
SlinegradientenergytheofSlopelossEddy
eee
ee
e
e
ωVω
Vω.cos ω
V(=Q/A)
Module 6
,,
.)(
exp...:sin
12
equationmomentumNowcasestheallinvealwaysisKthatEnsure
VVObviouslytcoefficienansionorncontractioKHere
SdxAFlosseddythegcauforceDrag
e
e
ee
+
>>
=
−= γ
AdVVdVdtdF
sccv
......
ρρ ∫∫∫∫∫∑ +∀=
as vdm
vmd==
)(β
( ) AdVVAdVVdxAVdtdF
outletinlet
......... ρρρ ∫∫∫∫∑ ++=
Module 6
Module 6
( )
( )dxqVQVdxqQand
directionvealongvelocityofcomponentbetoVConsider
factorcorrectionmomentumHere
AdVVAdVVdxAVdtdF
x
x
outletinlet
.......,.
)(
)1(.........
11
1
ρβρβ
β
ρρρ
==
+
=
++= ∫∫∫∫∑
)..(1 QVρβ ( ) )....(1 dxQVx
QV ρρβ∂∂
+
Forces acting on the C.V. (Contd..)
Body force
Friction force
Contraction/Expansion
forceWind force
Hydrostatic force
Boundary Reaction
Module 6
( ) ( )( ) ( )
( ) ( )
:exp,
........
..............
),1(,
1
111
forcetotaltheoutfindtoansionstheallcombiningNow
dxQVx
dxqVdxAVt
dxQVx
QVdxqVQVdxAVdtd
equationNow
x
x
∂∂
+−∂∂
=
∂∂
+++−
ρρβρ
ρρβρβρβρ
( )
( ) dxdxby
dxdxbyAdx
xyBdxWgAdxSgAdxSgAdxS
y
y
fef
ηηγ
ηηγγρρρρ
∂∂
−+
∂∂
−−∂∂
−−−−
∫
∫
0
00
Module 6
( ) ( )
( )
( )
( )
∂∂
+−
∂∂
=∂∂
−−
∂∂
+−∂∂
=∂∂
−−
==
∂∂
+−∂∂
=
∂∂
+−∂∂
=
∂∂
+−∂∂
=
∂∂
−−−−
AQ
xgAgAqV
tQ
gAxySSor
gAbySHRSHLdividingNowA
Qx
qVtQ
xygASSgA
SWgAssuA
Qx
qVtQ
QVx
qVtQ
dxQVx
qVAVt
AxygBWgASgASgASor
dxbySHRSHLdividingNow
xf
xf
ef
x
x
x
fef
211
0
2
110
2
11
11
1
0
.1
:""..&..,
.
,0&0min
.
..
...
:""..&..,
ββ
ββ
ββ
ββ
β
ρ
Module 6
( )
,""..&..
011,10
:)(
0
2
1
gbySHRSHLgMultiplyin
SSxy
AQ
xgAtQ
gAandqIf
channelsprismaticforequationmomentumofformonConservati
f =−−∂∂
−
∂∂
+
∂∂
== β
Local acceleration
term
Convective acceleration
term
Pressure force term
Gravity force term
Friction force term
0)(11 2
=−−∂∂
+
∂∂
+∂∂
fo SSgxyg
AQ
xAtQ
A
Module 6
( )
0)(.2.
0,,
0)(.2.
0)(
.11..21..1
0)(.1.1
.,
0)(11
:
2
tan
tan
2
2
tan
2
22
2
=−−∂∂
+
∂∂
∂∂
+
∂∂
+
∂∂
∂∂
+∂∂
=∂∂
=−−∂∂
+
∂∂
∂∂
+∂∂
+
∂∂
+
∂∂
∂∂
+∂∂
=−−∂∂
+
∂∂
∂∂
+
∂∂
+
∂∂
+
∂∂
+∂∂
=−−∂∂
+
∂∂
+∂
∂
=
=−−∂∂
+
∂∂
+∂∂
=
=
=
fo
tconsy
fotconsy
fo
tconsy
fo
fo
SSgxyg
xy
yA
AV
xVV
ty
yA
AV
tVor
xAchannelprismaticforNow
SSgxyg
xy
yA
xA
AV
xVV
ty
yA
AV
tVor
SSgxyg
xy
yAV
AxAV
AxVVA
AtAV
tVA
Aor
SSgxyg
AVA
xAtVA
Aor
VAQTaking
SSgxyg
AQ
xAtQ
A
EquationMomentumofformonConservati
Kinematic Wave
Diffusion Wave
Dynamic WaveModule 6
0)(
0)(..
2
2
=−−∂∂
+
∂∂
+
∂∂
+
∂∂
+
∂∂
+∂∂
=−−∂∂
+
∂∂
∂∂
+
∂∂
+
∂∂
+
∂∂
+∂∂
fo
fo
SSgxyg
xy
yV
xVV
xVV
ty
yV
tVor
SSgxyg
xy
yA
AV
xVV
xVV
tyB
AV
tVor
0)(.. =−−∂∂
+
∂∂
+
∂∂
+∂∂
+∂∂
+∂∂
fo SSgxyg
xVV
xyV
xVy
ty
yV
tV
Continuity Equation forNon-Conservation form(=0)
0)( =−−∂∂
+∂∂
+∂∂
fo SSgxyg
xVV
tV
Module 6
gAqVSS
xy
xV
gV
tV
g
gbySHRSHLDividingA
qVSSgxyg
xVV
tV
flowernalisthereifBut
xfo
xfo
1
1
)(1,""..&..
)(
,int
β
β
=−−∂∂
+∂∂
+∂∂
=−−∂∂
+∂∂
+∂∂
fo SSxy
xV
gV
tV
g=+
∂∂
−∂∂
−∂∂
−1
Steady, uniform flow
Steady, non-uniform flow
Unsteady, non-uniform flow
)24.6()(1
)23.6()(
)22.6(0)(
)(
)0(
)0(
1
1
1
1
gAqVSS
xy
xV
gV
tV
g
gAqVSS
xy
xV
gV
gAqVSSor
gAqVSS
flowuniformfortermonacceleraticonvectivexV
flowsteadyfortermonacceleratilocaltVHere
xfo
xfo
xfo
xfo
β
β
β
β
=−−∂∂
+∂∂
+∂∂
=−−∂∂
+∂∂
=+−
=−−
=⇒∂∂
=⇒∂∂
,flowuniformsteadyFor −
,flowuniformnonsteadyFor −
,flowuniformnonunsteadyFor −
Module 6
Lecture 4: Derivation of momentum equation (contd.)
Module 6
2D Saint Venant Equations
Obtained from Reynolds Navier-Stokes equations by depth-averaging.
Suitable for flow over a dyke, through the breach, over the floodplain.
Assumptions: hydrostatic pressure distribution, small channel slope
Module 6
.)()()(
.)()()(
.0)()(
31
222
2
31
222
2
eqmomentumyh
vuvgnyzgh
yhgh
yhv
xhuv
thv
eqmomentumxh
vuugnxzgh
xhgh
yhuv
xhu
thu
eqcontinuityyhv
xhu
th
b
b
+−
∂∂
−=∂∂
+∂
∂+
∂∂
+∂
∂
+−
∂∂
−=∂∂
+∂
∂+
∂∂
+∂
∂
=∂
∂+
∂∂
+∂∂
Solutions to St. Venant equations
Method of characteristics
Finite Difference methods
Explicit Implicit
Finite Element Methods
Solutions to St. Venant equations
Further reading - Note: NPTEL (Computational Hydraulics), http://www.efm.leeds.ac.uk/CIVE/CIVE3400/stvenant.pdf.
Module 6
Solutions to St. Venant equations Contd…
• Analytical
– Solved by integrating partial differential equations
– Applicable to only a few special simple cases of kinematic waves
• Numerical
– Finite difference approximation
– Calculations are performed on a grid placed over the (x, t) plane
– Flow and water surface elevation are obtained for incremental time
and distances along the channel
Module 6
x-t plane for finite differences calculations
Finite Difference Scheme (FDS)
Ups
trea
m b
ound
ary
Dow
nstr
eam
bou
ndar
y
Tim
e, t
Distance, x
Module 6
∆x ∆x
∆t h0, Q0, t0 h1, Q1, t0 h2, Q2, t0
h0, Q0, t1 h1, Q1, t1 h2, Q2, t2
∆x ∆x
∆t
i, ji-1, j i+1, j
i-1, j+1 i-1, j+1 i+1, j+1
x-t plane
Cross-sectional viewin x-t plane
Module 6
Finite Difference Scheme (FDS) Contd…
Module 6
Explicit Implicit
Temporal derivative
Spatial derivative
Temporal derivative
Spatial derivative
j ji iu uu
t t
+ −∂≈
∂ ∆
1
j ji iu uu
x x+ −−∂
≈∂ ∆
1 1
2
tuuuu
tu j
ij
ij
ij
i
∆−−+
≈∂∂ +
++
+
21
11
1
xuu
xuu
xu j
ij
ij
ij
i
∆−
−+∆−
≈∂∂ +
+++ 1
111 )1( θθ
Time step j+1
Time step j
Time step j-1
Space
Time
Reach
Spatial derivative is written using terms on known time line ‘j’
Explicit FDS
Module 6
Time step j+1/2
Time step j
Time Time step j+1
i i+1 i-1
Space
h1h3
h5
h7
Q
Q
Q
Centerpoint
Spatial and temporal derivatives use unknown time lines for computation ‘j+1’
Implicit FDS
Module 6
24
6
v2
v1
P1
P2
W
θ
Wsinθ
Rf
L
fs RPWPF −−+=∑ 21 sinθ
Momentum Analysis in an Open Channel
For a constant mass and a per unit width consideration in a rectangular channel,
( ) ( )12 vvqmvs −=∆ ρ
Module 6
Here,
Rf is the frictional resistance.
P1 and P2 are pressure forces per unit width given by:
Combining terms we get:
Considering a short section so that Rf is negligible and the channel slope is
small so that sinθ ≅ 0, the equation can be written as:
2yP
2γ=
)(sin22 12
22
21 vvqRWyy
f −=−+− ρθγγ
2
22
1
21 qv
2yqv
2y
ρ+γ
=ρ+γ
Momentum Analysis in an Open Channel Contd…
Module 6
Specific force plus momentum curve
• i.e, Mg
qv2y
gqv
2y 2
221
21 =+=+
M is the specific force plus
momentum and is constant for
both y1 & y2.
There are two possible depths
for a given M called sequent
depths. The depth associated
with the minimum M is yc.
MMc
y
yc
M
y1
y2
q1
q2
y = yc
Momentum Analysis in an Open Channel Contd…
Module 6
M is called the momentum function
or the specific force plus
momentum.
For a constant q, M can be plotted
against depth to create a curve
similar to the specific energy curve.
Under steady conditions, M is
constant from point to point along a
channel reach.
When q = q1 = q2 (at steady conditions when depth and velocity remain constant at a point)
v2
v1
y1
y2
12
Mg
qv2y
gqv
2y 2
221
21 =+=+
Momentum Analysis in an Open Channel Contd…
Module 6
Lecture 5: Energy equation and numerical problems
Module 6
Energy in Gradually VariedOpen channel flow
Module 6
In a closed conduit there can be a pressure gradient that drives the flow. An open channel has atmospheric pressure at the surface. The HGL (Hydraulic Gradient Line) is thus the same as the fluid surface.
Energy equation applied to open channel
p1 and p2 : pressure forces per unit width at sections 1 & 2 respectivelyv1 and v2 : velocity of flow at sections 1 & 2 respectively
and : energy coefficientsz1 and z2: elevations of channel bottom at sections 1 & 2 respectively w.r.t any datumhL: energy head loss of flow through the channel
1α 2α
Lhzg
vpzg
vp+++=++ 2
22
22
1
21
11
22α
γα
γ
Module 6
Simplifications made to the Energy Equation:1. Assume turbulent flow (α = 1).
2. Assume the slope is zero locally, so that z1 = z2
3. Write pressure in terms of depth (y = p / γ).
4. Assume friction is negligible (hL = 0).
i.e.
gvyg
vy 2222
2
21
1 +=+
21 EE =
Module 6
Energy equation applied to open channel Contd…
Specific Energy Equation
Module 6
Energy at a particular point in the channel = Potential Energy + Kinetic Energy
where y is the depth of flow, v is the velocity, Q is the discharge, A is the cross-
sectional flow area and E is the specific energy i.e energy w.r.t channel bottom.
2
2
2
2
2
gAQyE
org
vyE
+=
+=
Total energy
When energy is measured with respect to another fixed datum , it’s
called Total Energy
where z is the height of the channel bottom from the datum
Pressure head (y) is the ratio of pressure and the specific weight of
water
Elevation head or the datum head (z) is the height of the section
under consideration above a datum
Velocity head (v2/2g) is due to the average velocity of flow in that
vertical sectionModule 6
gvzyE 2
2++=
Example Problem
Module 6
Channel width (rectangular) = 2m, Depth = 1m, Q = 3.0 m3/s, Height above
datum = 2m. Compute specific and total energy
Ans: A = b*y = 2.0*1.0 = 2 m2
Specific energy =
Total energy =
Datum height + specific energy = 2.0 + 1.20 = 3.20 m
2
2
2gAQyE +=
2
2
2*81.9*231+=E
Specific Energy Diagram
Module 6
The specific energy can be plotted graphically as a function of depth of flow : E = Es + Ekwhere
)(2 2
2
energyKineticgAQEk =
)( energyStaticyEs =
2
2
2gAQyE +=
Specific Energy Diagram Contd…
Module 6
As the depth of flow increases,
the static energy increases and
the kinetic energy decreases,
The total energy curve
approaches the static energy
curve for high depths and the
kinetic energy curve for small
depths
ks EyEy 1∞⇒∞
Specific Energy Diagram Contd…
Module 6
As discharge (Q) increases, the specific energy curves move to the upper right
portion of the graph
Thus, for flat slope (+ other assumptions…) we can graph y against E:
(Recall for given flow, E1 = E2 )
Curve for different, higher Q.
For given Q and E, usually have 2 allowed depths:Subcritical and supercritical flow.
Module 6
The specific energy is minimum (Emin) for a particular critical depth –
Depth
Froude’s number = 1.0. & velocity = Vc.
Emin only energy value with a singular depth!
Depths < critical depths – supercritical flow (Calm, tranquil flow)
Froude Number > 1.0. V > Vc.
Depths > critical depths – subcritical flow (Rapid flow, “whitewater”)
Froude Number < 1.0. V < Vc.
Example: Flow past a sluice gate
Module 6
Critical Depth and Froude Number
It can easily be shown that at ,
Module 6
At the turning point (the left-most point of
the blue curve), there is just one value of
y(E).
This point can be found from
The Froude number can be defined as:
(Recall that the Reynolds number is the ratio of acceleration to viscous forces).The Froude number is the ratio of acceleration to gravityPerhaps more illustrative is the fact that surface (gravity) waves move at a speed of
Flows with Fr < 1 move slower than gravity waves.
Flows with Fr > 1 move faster than gravity waves.
Flows with Fr = 1 move at the same speed as gravity waves.
Module 6
Critical Depth and Froude Number Contd…
Flows sometimes switch from supercritical to subcritical:
(The switch depends on upstream and downstream velocities)
Gravity waves: If you throw a rock into the water, the entire circular wave will travel downstream in supercritical flow.
In subcritical flow, the part of the wave trying to travel upstream will in fact move upstream (against the flow of the current).
Module 6
Critical Depth and Froude Number Contd…
Example Problem
Will the flow over a bump be supercritical or subcritical?
As it turns out:Left = subcriticalRight = supercritical
Using the Bernoulli equation for frictionless, steady, incompressible flow along a streamline:
or
Left
Left
Right
Right
Module 6
Apply Bernoulli equation along free surface streamline (p=0):
For a channel of rectangular cross-section,
Module 6
Example Problem Contd…
Substitute Q = V z b into Bernoulli equation:
To find the shape of the free surface, take the x-derivative:
Solve for dz / dx:
Module 6
Example Problem Contd…
Since subcritical: Fr < 1supercritical: Fr > 1
Subcritical flow with dh / dx > 1 dz / dx < 1Supercritical flow with dh / dx > 1 dz / dx > 1
if flow is subcritical if flow is supercritical
Module 6
Example Problem Contd…
Hydraulic Jump
Module 6
There is a lot of viscous dissipation ( = head loss ) within the hydraulic jump.
Module 6
Hydraulic Jump Contd…
Apply the momentum equation:
Momentum equation is used here as there is an unknown loss of energy (where mechanical energy is converted to heat).But as long as there is no friction along the base of the flow, there is no loss of momentum involved.
Module 6
Hydraulic Jump Contd…
Momentum balance:
The forces are hydrostatic forces on each end:
(where and are the pressures at centroids of A1 and A2 )
Module 6
Hydraulic Jump Contd…
If y1 and Q are given, then for rectangular channel
is the pressure at mid-depth.
Here, entire left-hand side is known, and we also know the first term on the
right-hand side. So we can find V2.
Module 6
Hydraulic Jump Contd…
1) A rectangular channel 4m wide has a flow discharge of 10.0 m3/s and depth of
flow as 2.5 m. Draw specific energy diagram and find critical and alternate
depth.
2) A triangular channel with side slopes having ratio of 1:1.5 has a discharge
capacity of 0.02 m3/s. Calculate:
a. critical depth
b. Emin
c. Plot specific energy curve
d. Determine energy for 0.25 ft and alternate depth
e. Velocity of flow and Froude number
f. Calculate required slopes if depths from d are to be normal
depths for given flow.
Exercises
Module 6
Highlights in the Module
Reynolds' transport theorem (Leibniz-Reynolds' transport theorem) is a 3-D generalization of the Leibniz integral rule.
Control volume is a definite volume specified in space. Matter in a control volume can change with time as matter enters and leaves its control surface.
Reynolds Transport Theorem states that the total rate of change of any extensive property B of a system occupying a control volume C.V. at time ‘t’ is equal to the sum of:
a) the temporal rate of change of B within the C.V.b) the net flux of B through the control surface C.S. that surrounds
the C.V.
Module 6
Highlights in the Module Contd…
St.Venant equations are derived from Navier-Stokes Equations for shallow water flow conditions. The solution of the St. Venant equations is known as dynamic routing,
which is generally the standard to which other methods are measured or compared.
Forces acting on the C.V. in an open channel flow are gravity force, friction force , contraction/expansion force, wind shear force and unbalanced pressure forces.
Solutions to St. Venant equations : Method of characteristics Finite Difference methods : Explicit, Implicit Finite Element Methods
Module 6
Hydrological Statistics
Prof. Subhankar KarmakarIIT Bombay
Module 77 Lectures
Objective of this module is to learn the fundamentalsof stochastic hydrological phenomena.
Module 7
Module 7
Topics to be covered
Statistical parameter estimation
Probability distribution
Goodness of fit
Concepts of probability weighted moments & l-moments
Lecture 1: Introduction to probability distribution
Module 7
Introduction
Deterministic model
– Variables involved are deterministic in nature
– No uncertainty for a given set of inputs
E.g. Newtonian model(2nd law of motion)
Probabilistic/Stochastic model
– Variables involved are random in nature.
– There is always uncertainty for a given set of inputs
E.g. Rainfall-runoff model
212s ut ft= +
In a deterministic model, a given input yields the same output always, while a probabilistic model yields different outputs for multiple trials with the same input.
Module 7
Random Variables
Random Variable (R.V.) is a variable whose value can not be predicted
with certainty before it actually takes over the value. R.V may be discrete or
continuous.
E.g. rainfall, stream flow, soil hydraulic properties
(permeability, porosity, etc.), evaporation, diffusion, temperature, ground
water level, etc.
Notations:– Capital letters X, Y, Z indicates R.V.– Small letter x, y, z indicates value of the R.V.
X x≤
Rainfall 30 mm (value)
If X is a R.V. & Z(X) is function of X ==> Z is a R.V.
E.g. Rainfall is R.V, so Runoff is R.V.
Module 7
Discrete random variables
If the set of values for a R.V. can be assumed to be finite (or
countable infinite), then RV is said to be a discrete RV.
E.g. No. of raining days in a month (0,1,2,…,31)
No. of particles emitted by a radio-active material. Continuous random variables
If the set of values for a R.V. can be assumed to be infinite,
then RV is said to be a continuous R.V.
E.g. Depth of rainfall in a period of given month
Random Variables Contd..…..
By using ‘class interval’ we can convert continuous random variables to discrete random variables.
Module 7
Probability Distribution
Discrete Variable
X1 X2 X3 Xn
Splices
P [X=xi] Probability mass function (pmf)
Important conditions:
(i) 0 ≤ P [X=xi] ≤ 1
(ii) P [X=xi] = 1
[Here 1 is likelihood of occurrence]
Cumulative distribution function (CDF)
CDF is a step function
i∑
x1 x2 x3 xn
1.0
x
p(x1)p(x1)+ p(x2) for xi ≤ x2
p(x1)+ p(x2) + p(x3)
NotationP [ X = x] = p(x)P
[ ] [ ]i
i ix x
P X x P X x≤
≤ = =∑{ }1 2( ) 0 , ,......, nP X x X x x x= = ∀ =
Cumulative distribution function (cdf)
Module 7
Probability Distribution Contd..…..
Continuous Variable Probability density function
(PDF), f(x) does not indicate
probability directly but it indicates
probability density
CDF is given by
f(x)
ax
+α
a x
F(x)
+α-α
1.0
( ) [ ] ( ) a
F a P X a f x dxα−
= ≤ = ∫
Module 7
-α
Piecewise Continuous Distribution
d x
f(x)
0
f1(x)
f2(x)
P[X=d]
spike
1 2 ( ) + P[X=d]+ ( ) = 1
Then, P[X<d] P[X d];
Simillarly, P[X>d] P[X d];
d
d
f x dx f x dxα
α−
≠ ≤
≠ ≥
∫ ∫
d x0
f(x)
P[X = 0]
X = 0Module 7
d x0
Jump ΔFP[X=d]X=d
CDF for this case
CDF within an interval
Δx
x 2 2 2
i i i
x x xx x∆ ∆ ∆ − +
xi
f(xi)
i
ii
ii
ii
xxf
xxxx
ervaltheincurvetheunderAreaxxFxxF
xxXxxP
∆=
∆
+∆
−
=∆
−−∆
+=
∆+≤≤
∆−
*)(2
,2
int)2
()2
(
]22
[
Here, interval is given by (x-∆x/2, x+∆x/2)
Module 7
Example Problem
Estimate the expected relative frequencies within each class of interval 0.25. 23( ) ; 0 x 5
125
xf x = ≤ ≤x1 x2 x3
0 0.25 0.75 1.25 1.75 5.0
xi f(xi)=3x2i/125 f(xi) x Δx = expected rel. freq.
0.25 0.0015 0.00075
0.75 0.0135 0.00675
1.25 0.0375 0.01875
1.75 0.0735 0.03675
2.25 0.1215 --
2.75 0.066 --
3.25 -- --
3.75 0.3375 0.16875
4.25 0.4335 0.21675
4.75 0.5415 0.27075
Sum 0.9975 ≈1.0
Module 7
Bivariate Random Variables
(X,Y) – two dimensional random vector or random variable
e.g. Temperature EvaporationTemperature Radiation Coeff.Rainfall RechargeRainfall Runoff
Variables may or may not be dependent
X = rainfall
y = runoff
X
Y
(X+Y)
Case 1: (X,Y) is a 2-D discrete random vector
The possible values of may be represented as (Xi, Yj), i=1,2,….n & J=1,2,….n
Case 2: (X,Y) is a 2-D continuous random variable
Here, (X,Y) can assume all possible values in some non-countable set.
Module 7
1. Discrete 2-D Random Vector
( ) [ ] ∑∑≤
∞−
≤
∞−=≤≤=
xX yY
YXpyYxXPyxF ),(,,
i=1 2 3 4 5 6 P[Y]
X Y 0 1 2 3 4 5
j=1 0 0 0.01 0.03 0.05 0.07 0.09 0.25 P[Y=0]
2 1 0.01 0.02 0.04 0.05 0.06 0.08 0.26
3 2 0.01 0.03 0.05 0.05 0.05 0.06 0.25
4 3 0.01 0.02 0.04 0.06 0.06 0.05 0.24 P[Y=3]
0.03 0.08 0.16 0.21 0.24 0.28
P[x=0] P[x=5]
( ) 1, =∞∞F
[ ]( )
1)8,7(1,)4,7(
35.02,3)2,3(
1),(6
1
4
1
==∞∞=
=≤≤=
=∑∑= =
FFF
YXPF
YXpi j
ji
E.g. Calculate the following using the values from the given table.
1) F(3,2)
2) F(7,4)
3) F(7,8)
Module 7
( , ) joint pdf of (X,Y)f(X,Y) 0
( , ) 1
F(x,y) joint cdf of (X,Y)
= P[X , ]
= ( , )
Probablity of hatched region on the plane
α α
α α
α α
− −
− −
→
≥
=
→
≤ ≤
∫ ∫
∫ ∫y x
If f X Y
f x y dxdy
x Y y
f x y dxdy
= ( , ) ∫B
f x y dB
B
f(x,y) plane
2. Continuous 2-D Random Vector
Module 7
F( , )=1
F(- ,y) = F(x, - ) = 0
α α
α α
Example Problem 1
Consider the flows in two adjacent streams. Denote it as a 2-D RV (X,Y), with a joint pdf,
( , ) if 5,000 x 10,000 and 4,000 y 9,000 Solution: T
= ≤ ≤≤ ≤
f X Y C
2
o get C,
( , ) 1 1
1 C =
(5000),
P(B) = ( , )
α α α α
α α α α− − − −
= ⇒ =
∴
∫ ∫ ∫ ∫
∫ ∫B
f x y dxdy C dxdy
For a watershed region B
f x y dxdy
Module 7
9000
5000 5000
9000 9000
5000 5000 5000
90002
5000
etermine P[X Y]
= 1- P[X Y] = 1 - ( , )
= 1 - = 1 - [ 5000]
= 1- C 50002
=
≥
≤
−
−
∫ ∫
∫ ∫ ∫
y
y
D
f x y dxdy
Cdxdy y dy
y y
90002 2
5000
(9000) (5000) 1- C 45000000 250000002 2
17 = 25
− − +
Module 7
Example Problem 1 Contd…
2
2 2
2 2 2 2
2
1F(x,y) = ( , ) = (5000)
4000 = - dx(5000) (5000)
4000 (5000) 5000 4000 = -(5000) (5000) (5000) (5000)
4 = -(5000) 5 (50
α α α α
α
− − − −
−
×− +
×
∫ ∫ ∫ ∫
∫
y yx x
x
f x y dydx dydx
y
yx yx
xy x 4 (Ans.)500) (5000)
− +y
2
F(10000,9000)
10000 9000 4(10000) 9000 4= - 5(5000) 5 (5000) (5000)
= 3.6 - 1.6 -1.8 +0.8
=
1.0
×− +
×
Module 7
Calculate F(10000,9000):
Example Problem 1 Contd…
2
12 1 2 3 22
0 0 0 0
f(x,y) x +xy 0 x 1 & 0 y 2 = 0 elsewhere
Find, P[x+y 1], verify f(x,y)dxdy = 1
Solution :x x yxy(x )dxdy= dy 3 3 6
α α
α α− −
= ≤ ≤ ≤ ≤
≥
+ +
∫ ∫
∫ ∫ ∫22 2
0 0
1 y 1 y 2 4= dy= = 3 6 3 12 3 12
= 1 [verified]
+ + + ∫
Module 7
Example Problem 2
1 12
0 0
11 22
0 0
1 22
0
1 2 3 2 3
0
[ 1]
1 [ 1]
= 1- 3
=1- dx6
(1 2 )= 1- (1 ) dx6
6 6 2 )= 1- dx6
x
x
P X Y
P X Y
xyx dydx
y xx y
x x xx x
x x x x x
−
−
+ ≥
= − + ≤
+
+
− +− +
− + − +
∫ ∫
∫
∫
∫
12 3
0
13 4 2
0
1=1- (4 5 ) dx6
1 4 5=1-6 3 4 2
1 4 5 116 3 4 2
1 716 12
6572
x x x
x x x
− +
− +
= − − +
= −
=
∫
Module 7
Example Problem 2 Contd…
Marginal Distribution Function
1
1
Marginal Distribution of X is [ ] ( , ), i
Marginal Distribution of y is [ ] ( , ),
α
α=
=
= ∀
= ∀
∑
∑
i i jj
j i ji
p x p x y
q y p x y j
α
αα
α
−
−
∫
∫
Marginal distribution function of x is given by g(x) = ( , )
of y is given by h(y) = ( , )
[g(x) & h(y) are also pdf and should satisfy the conditions]
f x y dy
and f x y dx
Discrete variables
Continuous variables
Module 7
Marginal Distribution Function Contd…
α
α
α α
−
≤ ≤ = ≤ ≤ − ≤ ≤
= ×
∫ ∫
∫
[ ] P[c , ]
= ( , )
= ( )
[ ( , ) ( ) ( ) ......stochastically independent
( ) is original distribution of the 1-D ran
d
c
d
c
P c X d X d y
f x y dy dx
g x dx
f x y g x h y
g x
α−∫
dom variable X
Remember C.D.F = F(x) = ( )
( ) is same as f(x), i.e. orginal density function of x.
x
g x dx
g x
Module 7
= − − ≤ ≤ ≤ ≤
∴ ≤ ≤ − −
−
∫ ∫
,
2 2
0 02
Function P ( , ) (5 ( ) ) 0 x 2 and 0 y 2 can serve as a 2
bivariate contnuous probablity density function.
i) Find the value of C
Probability (x 2 & y 2) = (5 ( ) ) dsdt=1 2
54
x yyx y C x for
SC t
sC s [ ] − = ⇒ − − =
∫ ∫
22 2
0 00
1 10 1 2 1ts dt C t dt
[ ]− = ⇒ − =
− = ⇒ =
∫2
2 20
0
9 2 1 [9 ] 1
1[18 4] 1 14
C t dt C t t
C C
Module 7
Example Problem
= ≤ ≤
− −
− − − −
∫ ∫
∫
,
0 0
2 2 2
0
P ( , ) ( , )
= (5 ( ) ) /14 dsdt 2
1 1 = [5 ] = 514 4 14 4 2
x y
yx
x
x y prob X x Y y
S t
y xy x yy ty dt xy − −
1 1 1 = 5 = 0.30414 4 2
≤ ≤ii) probability of x 1 & y 1 Find
iii) "marginal desities" P (x) & P (Y) x yFind
=
− −
− −
− − −
∫
∫
2
,02
022
0
P (y) ( , )
1 = (5 ) 2141 = 5
14 2 21 1 = [10 2]= (8 )
14 14
y X YP t y dt
y t dt
yt tt
y y
= − −
− −
− −
−
∫ ∫2 2
,0 0
22
0
P (x) ( , ) = (5 ) /14 2
1 = 514 41 = [10 1 2 ]
141 = (9 2 )
14
x X YsP x s ds x ds
ss xs
x
x
Module 7
α α
α −∴ = ≤
− −
∫ ∫
, ,
, ,0 0
2 2
0
iv) cumulative marginal distributions P (x, ) & Px ( ,Y)
9 2P (x, ) ( )= ( ) = ( ) 14
9 9 = = 14 14
x y x y
x x
x y X
x
Find
xprob X x P t dt dx
x x x x
α − −
= ≤
−=
−=
∫ ∫2
,0 0
2
2
8 8 / 2P ( , ) ( )= ( ) = ( ) = 14 14
9 Check putting x=2 , 114
16& y=1, 128
y y
x y yy y yy prob Y y P s ds ds
x x
y y
− −
−− −
−
/ ,
/ ,
v) "conditional densities" (5 )2 P (x/y) = Px (x,y)/P (y) =
(8 )(5 )2P (y/x) = Px (x,y)/P (x) =
(9 2 )
x y x y y
x y x y x
Findy x
yy x
andx
Module 7
Conditional Density Functions
Let (X,Y) is a 2-D random vector, with joint density function of f(x,y). Let g(x)
and h(y) be the marginal density functions of X and Y respectively.
The CDF of X, given Y=y is defined as,
and the CDF of Y, given X=x is defined as,
( , )( | ) but h(y)>0( )
f x yg x yh y
=
( , )( | ) but g(x)>0( )
f x yh y xg x
=
α α α
α α αα
α
− − −
−
≥
= =
∴ =
∫ ∫ ∫
∫
( | ) 0 as f(x,y) and h(y) are (+ve)( , ) 1 1( | ) 1 = ( , ) = ( ) = 1( ) ( ) ( )
( , ) ( )
g x yf x yg x y dx dx f x y dx h yh y h y h y
f x y dx h y
Module 7
Conditional Density Functions Contd….
CDF = G(x|y) = P[ X x|Y = y ] = ( | )
y belongs to certain region, i.e. y R (R is a region)
( , )( | )
( )
Now cumulative conditional distribution function
= P [ X x | y R]
α
α−
≤
∈
∴ ∈ =
≤ ∈
∫
∫∫R
R
g x y dx
where
f x y dyg x y R
h y dy
= F( x | y R)
= F(X | y R) ( | )α−
∈
∈ = ∈∫x
g x y R dx
Module 7
Independence of random variables
( | ) ( ), when X & Y are statistically independent variables
( , )g(x|y) ( )( )
f(x,y) = g(x) x h(y), then x, y are independent
=
= =
g x y g x
f x y g xh y
If
( )
( )( )
0 0
( , ) ; x>0; y>0
g(x) = = 1
= e , x>0 & h(y) = e , y>0
( , ) ( ) h(y) x &y are independent
αα
− +
− +− +
− −
=
∴ −
∴ = × ∴
∫
x y
x yx y
x y
f x y eee dy
f x y g x
For e.g.,
Module 7
Functions of random variables
− +2
1:: variable, p(x) = f(x) = , x=2,3,4,5, y= x 7 12
Examplecx discrete xx
7 takes two values 0 & 2, p(y=0) = p(x=3) + p(x=4)= 3 4 12
7p(y=2) = p(x=2) + p(x=5) = 2 5 10
+ =
+ =
c c cy
c c c
x 2 3 4 5y 2 0 0 2
p(x) c/2 c/3 c/4 c/5
Simultaneous occurrence Joint density function Distribution of one variable irrespective of the value of the other
variables Marginal density function Distribution of one variable conditioned on the other variable
Conditioned distribution
Module 7
Functions of Random variables Contd…..
2
0
2 :
functions of RVs:
For continuous curve
f(x) = e ; x>0
y = 2x+11P[y 5] = P [ X ]
2
= P[X 2]
= 1- P[X 2]
= 1-
−
−
−≥ ≥
≥
≤
∫
x
x
Example
for
y
e dx
2
0
2 0
2
2
2 2
= 1-
= 1- [ ]1 1
= (Ans.)
(Ans.)1
x
xx
e dx
e e
e
Alternatively
ee dx eαα
−
− −
−
−− −
−− −
=
= = −
∫
∫Module 7
2
i) g(x|y) = f(x,y) / h(y)
1 14 (10 2 ) = (5 ) = , 0 x 2, 0 y 2214 (8 ) 2(8 )
Conditional CDF:
(10 )( | ) [ | ]= ( | ) = 2(8 )α−
− −− − × ≤ ≤ ≤ ≤
− −
− −= ≤ =
−∫x
y xy xy y
x xy xg x y P X x Y y g x y dxy
≤ =
− − − −=
−
2
3, [ 1, ]2
3 3(10(1) (1)( ) 1 ) 10 1 152 23 = a[1| ] = = 2 3 13 262(8 ) 2( )2 2
Now P X Y
≤ ≤ ≤ ≤ ≤
− −
−
∫ ∫
∫ ∫
1 1
0 01 1
ii) Now, g(x|y 1] 0 x 2; 0 y 2
1( , ) (5 )14 2 = =
( ) (8 ) /14o o
yf x y dy x dy
h y dy y dy
Module 7
− − − − − − − − −
−
∫
∫
1 12
0 01 12
0
(5 ) 5 152 4 4 = = = 18(8 ) 8 22
(19 4 ) = 30
o
y yx dy y xy x
yy dy y
x
[ ]2
0
2
1 x= 2
19 4 19) | 1 = 30 30 15
,
18 21 1| 1 | 1 = 2 2 15 15
18 2 17 = 2 15 15(4) 30
− ≤ = −
≤ ≤ = < −
− =×
∫x
for
x x xiii F x y dx
Now
x xP X Y F Y
x
Module 7
Properties of RV
Population: A complete assemblage of all possible RV
Sample: A subset of population
Realization: A time series of the RVs actually realized. They are continuous.
All samples are not realizations.
Observation: A particular value of the RV in the realization
Note: ARMA is a synthetically generated realization
Module 7
Lecture 2: Statistical Moments
Module 7
Introduction
Measures of Central Tendency:
Mean,
Arithmetic average (for sample),
Mode,
Median
Measures of Spread or Distribution:
Range [(xmax-xmin)],
Relative Range [=(range/mean)],
Variance,
Standard deviation,
Coefficient of variationModule 7
α
α
α
α
α
α
µ
µ
µ
µ
+
−
+
−
+
−
=
= =
=
=
∫
∫
∫
0
0 00
01
n moment of the area ( )
moment of the area ( ) 1
Expected value of x is defined as:E(x) = first moment about the origin
= ( ) ( )
mo
th nn
th
th
x f x dx
o x f x dx
x f x dx simply
nα
α
µ+
−
= −∫ment about the expected value ( ) ( ) ...for population (not sample)nn x u f x dx
Moments of Distribution
dxx
f(x)
+α-α
x
Module 7
Moments of Distribution Contd…
[ ] [ ]
[ ] [ ] [ ]1 2 1 2
Therefore,
i) E(x) = ( )
ii) Expected value of a function of RV
= E[g(x)]
= ( ) ( )
iii) E(C) = C
iv) E C g(x) = C E g(x)
v) E g (x) g (x) E g (x) E g (x)
... "Expe
x f x dx
g x f x dx
α
α
α
α
µ+
−
+
−
=
± = ±
∫
∫
ctations" is an operator for population [- , + ]α α
dx x
f(x)
+α-α
xf(x)
μ = E(x)
Module 7
2
n
2
V is wind velocity with a pdf f(v) = 1/10; 0 v 10. The pressure at point isgiven by = 0.003 v . Find the expected value of pressure.
Sol :1 f(v) = ; 0 v 10; = 0.003 v
10
E( ) =
≤ ≤ ωω
≤ ≤ ω
ω
2
2
?
E( ) = g( )d , given = 0.003 v or d = 0.006 vdv
1given, f(v) = & = 0.003 v g( )will be monotonically increasing function.10
dv 1 0.0274Therefore, g( ) = f(v) d 10
=
+α
−α
ω ω ω ω ω ω
ω ∴ ω
ω = ×ω ω
∫
1 22 0.00274( ) as = 0.003 v−ω ω
Module 7
Example Problem
12
1 12 2
0.330.3 21
2
00
1or v = 18.257 or 18.257 ( )0.003 2
1 1or g(w) = 18.257 ( ) = 0.9129 (w)10 2
Now, E(w) = 0.9129 (w) = 0.9129 3
2
2 = 0.9129 [0.3]3
−
− −
−
= = × ×
× × ×
× ×
× ×
∫
dvw w wdw
w
ww dw
32
0 10 0 18.257 10
0 w 0.3
= 0.1 (Ans.)
≤ ≤ ≤ ≤ ≤ ≤
vor wor
Module 7
Example Problem Contd…
Probability weighted moments
Given a random variable X with a cumulative distribution function F, the probability weighted moments are defined to be:
Two special cases are:
For an ordered sample x1:n <= x2:n <= ... <= xn:n, unbiased estimators of
n
r j :nj
n
r j :nj
( n j )(n j )...(n j )a xn (n )(n )...(n r )
( j )( j )...( j r )b xn (n )(n )...(n r )
=
=
− − − − +=
− − −− − −
=− − −
∑
∑
1
1
1 1 11 2
1 1 21 2
p r sM( p,r ,s ) E[ X { F( x )} { F( x )} ]= −1
rr M( , ,r ) E[ X{ F( x )} ]α = = −1 0 1
rr M( ,r , ) E[ X{ F( x )} ]β = =1 0
Module 7
L-moments
L-moments differ from conventional moments in that they are calculated using
linear combinations of the ordered data.
Population L-moments
For a random variable X, the rth population L-moment is
where Xk:n denotes the kth order statistic (kth smallest value) in
an independent sample of size n from the distribution of X and denotes expected
value.
( )rrk
r r k:rkkr ( ) EXλ
−−−
−=
= −∑11
1
01
Module 7
L-moments Contd…
In particular, the first four population L-moments are
The first two of these L-moments have conventional names:
( )( )( )
: :
: : :
: : : :
EXEX EX /
EX EX EX /
EX EX EX EX /
λλ
λ
λ
=
= −
= − +
= − + −
1
2 2 2 1 2
3 3 3 2 3 1 3
4 4 4 3 4 2 4 1 4
22 33 3 4
mean, L mean or L location,L scale
λλ
= − −= −
1
2
Module 7
L-moments Contd…
Sample L-moments
Direct estimators for the first four L-moments in a finite sample of n observations are
where xi is the ith order statistic and is a binomial coefficient.
( )( ) ( ) ( ){ }( ) ( ) ( ) ( ) ( ){ }( ) ( ) ( ) ( ) ( ) ( ) ( ){ }
n n
ii
n i n in
ii
n i i n i n in
ii
n i i n i i n i n in
ii
x
x
x
x
−
=
− − −
=
− − − − −
=
− − − − − − −
=
=
= −
= − +
= − + −
∑
∑
∑
∑
1
1 1 1
1 1
1 2 1 11
1 1 1
3 3 2 1 1 21
1 1 1 1
3 4 3 2 1 1 2 31
121 231 3 34
Module 7
L-moments Contd…
L-moment ratios
A set of L-moment ratios, or scaled L-moments, is defined by
The most useful of these are τ3 ,called the L-skewness, and τ4, the L-kurtosis.r r / ,r , ...τ λ λ= =2 3 4
There are two common ways that L-moments are used:
As summary statistics for data.
To derive estimates for the parameters of probability distributions.
Module 7
Lecture 3: Measures of central tendency and dispersion
Module 7
Measures of Central Tendency
Module 7
1
1
E(x) = = ( ) , for parameter estimate
Arithmatic Mean ( ) = .......for sample estimate
Most frequently occuring value
1) 0 & 0
2) Value of x associated
Mean:
Mo
de:
with ma
α
α
µ+
−
=
=
∂ ∂= <
∂ ∂
∫
∑n
ii
n
i
x f x dx
x x
f fx x
f ( )
The pdf takes the maximum value at the mode.
x x ix
Measures of Central Tendency Contd…
Module 7
It divides the area under the pdf curve into two halves.
i.e Area is 50%
= ( ) = P[X ] = 0.5
:
[ It is the observation such that
half the values lie on either side of it ]
Median:
µ
µ µ−
≤∫med
med x medd
n
P x dx
Def
(b)
Median
Median
(c)
Median
(a)
[ ]
[ ] [ ]
[ ] [ ] [ ]
α α
α αα α α α
α α α α
− −
− − − −
+ = +
+ +
∴ + = +
∫ ∫
∫ ∫ ∫ ∫
L , E ( ) ( , )
= ( , ) y ( , ) =E E
E E E
et X Y x y f x y dxdy
x f x y dxdy f x y dxdy X Y
X Y X Y
[ ]α α
α α
α
α
− −
−
= ∫ ∫
∫
(X, Y) an independent RVs with a joint pdf of f(x,y)
E , ( , )
For independent variable f(x,y) = g(x) ( ) ( x and y are indpendent)
= (
X Y xy f x y dxdy
x h y because
xy g x [ ] [ ]
[ ] [ ] [ ]
α α α
α α α− − −
• +
∴ + = ×
∫ ∫ ∫) ( ) = ( ) y h( ) = E E
E E E (If x and Y are independent)
h y dxdy x g x dx y dy X Y
X Y X Y
Module 7
Measures of Central Tendency Contd…
Measure of “Spread” or “Dispersion”
Range : (Max - min) value
α
α
µ
µ
σ
−
−∫
2
2
2
Most important measure of dispersion
= Second moment about
Variance:
the mean
= ( ) ( )
=
x f x dx
= E(X)
= Expected value of x
Module 7
Measure of “Spread” or “Dispersion” Contd…
( )
[ ]
[ ]
[ ]
2 2
2
2 2
2 2
2 2
2 2 2
2 2
2
= ( ) ( )
= E
= E 2
= E 2E E
= E 2 E
= E 2
= E
= E E( )
x f x dx
X
X X
X X
X X
X
X
X X
α
α
σ µ
µ
µ µ
µ µ
µ µ
µ µ
µ
−
−
−
− +
− +
− +
− +
−
−
∫
2 2σ=
Module 7
( )( )
=
−=
−
∑2
2 1
2
2
Sample Estimate, s1
= + positive square root of variance
Co-effcient of Variation:
=
CoV = Cv =
i) Var(C) = 0ii) Var(Cx) = C var(x)iii) Var(a + bx
Standard Dev
) =
iation
:
b
n
ii
x x
n
Sx
σ σ
σµ
2 var( )x
Lecture 4: Introduction to unit hydrograph
Module 3
Unit hydrograph (UH)
• The unit hydrograph is the unit pulse response function of a linear hydrologicsystem.
• First proposed by Sherman (1932), the unit hydrograph (originally namedunit-graph) of a watershed is defined as a direct runoff hydrograph (DRH)resulting from 1 in (usually taken as 1 cm in SI units) of excess rainfallgenerated uniformly over the drainage area at a constant rate for an effectiveduration.
• Sherman originally used the word “unit” to denote a unit of time. But sincethat time it has often been interpreted as a unit depth of excess rainfall.
• Sherman classified runoff into surface runoff and groundwater runoff anddefined the unit hydrograph for use only with surface runoff.
Module 3
The unit hydrograph is a simple linear model that can be used to derive the
hydrograph resulting from any amount of excess rainfall. The following basic
assumptions are inherent in this model;
1. Rainfall excess of equal duration are assumed to produce hydrographs
with equivalent time bases regardless of the intensity of the rain
2. Direct runoff ordinates for a storm of given duration are assumed directly
proportional to rainfall excess volumes.
3. The time distribution of direct runoff is assumed independent of
antecedent precipitation
4. Rainfall distribution is assumed to be the same for all storms of equal
duration, both spatially and temporally
Unit hydrograph Contd….
Module 3
Terminologies1. Duration of effective rainfall : the time
from start to finish of effective rainfall
2. Lag time (L or tp): the time from the
center of mass of rainfall excess to the
peak of the hydrograph
3. Time of rise (TR): the time from the start
of rainfall excess to the peak of the
hydrograph
4. Time base (Tb): the total duration of the
DRO hydrographBase flow
Direct runoff
Inflection point
TRtp
Effective rainfall/excess rainfall
Q (c
fs)
Module 3
Derivation of UH : Gauged watershed
1. Storms should be selected with a simple structure with relatively uniform spatial
and temporal distributions
2. Watershed sizes should generally fall between 1.0 and 100 mi2 in modern
watershed analysis
3. Direct runoff should range 0.5 to 2 in.
4. Duration of rainfall excess D should be approximately 25% to 30% of lag time tp
5. A number of storms of similar duration should be analyzed to obtain an average
UH for that duration
6. Step 5 should be repeated for several rainfall of different durations
Module 3
Unit hydrograph
Rules to be observed in developing UH from gaged watersheds
1. Analyze the hydrograph and separate base flow
2. Measure the total volume of DRO under the hydrograph and convert time to
inches (mm) over the watershed
3. Convert total rainfall to rainfall excess through infiltration methods, such that
rainfall excess = DRO, and evaluate duration D of the rainfall excess that
produced the DRO hydrograph
4. Divide the ordinates of the DRO hydrograph by the volume in inches (mm)
and plot these results as the UH for the basin. Time base Tb is assumed
constant for storms of equal duration and thus it will not change
5. Check the volume of the UH to make sure it is 1.0 in.(1.0mm), and graphically
adjust ordinates as required
Module 3
Unit hydrograph
Essential steps for developing UH from single storm hydrograph
Obtain a Unit Hydrograph for a basin of 315 km2 of area using the rainfall and stream flow data tabulated below.
Time (hr) Observed hydrograph(m3/s)
0 100
1 100
2 300
3 700
4 1000
5 800
6 600
7 400
8 300
9 200
10 100
11 100
Time (hr)
Gross PPT(GRH) (cm/h)
0-1 0.51-2 2.52-3 2.53-4 0.5
Stream flow data Rainfall data
Module 3
Unit hydrograph
Example Problem
• Empirical unit hydrograph derivation separates the base flow from the observedstream flow hydrograph in order to obtain the direct runoff hydrograph (DRH). Forthis example, use the horizontal line method to separate the base flow. Fromobservation of the hydrograph data, the stream flow at the start of the rising limbof the hydrograph is 100 m3/s
• Compute the volume of direct runoff. This volume must be equal to the volume of the effective rainfall hyetograph (ERH)
VDRH = (200+600+900+700+500+300+200+100) m3/s (3600) s = 12'600,000 m3
• Express VDRH in equivalent units of depth:
VDRH in equivalent units of depth = VDRH/Abasin = 12'600,000 m3/(315000000 m2) = 0.04 m = 4 cm
Module 3
Unit hydrograph
Example Problem Contd…
Obtain a Unit Hydrograph by normalizing the DRH. Normalizing implies dividing theordinates of the DRH by the VDRH in equivalent units of depth
Time (hr) Observed hydrograph(m3/s)
Direct Runoff Hydrograph (DRH) (m3/s)
Unit Hydrograph (m3/s/cm)
0 100 0 01 100 0 02 300 200 503 700 600 1504 1000 900 2255 800 700 1756 600 500 1257 400 300 758 300 200 509 200 100 25
10 100 0 011 100 0 0
Module 3
Module 3
Unit hydrograph
Example Problem Contd…
0
200
400
600
800
1000
1200
0 2 4 6 8 10 12
Q (m
3 /s)
Time (hr)
Observed hydrograph
Unit hydrograph
DRH
• Determine the duration D of the ERH associated with the UH obtained in 4. In order to do this:
1. Determine the volume of losses, VLosses which is equal to the difference between the volume of gross rainfall, VGRH, and the volume of the direct runoff hydrograph, VDRH . VLosses = VGRH - VDRH = (0.5 + 2.5 + 2.5 +0.5) cm/h 1 h - 4 cm = 2 cm
2. Compute the f-index equal to the ratio of the volume of losses to the rainfall duration, tr. Thus, ø-index = VLosses/tr = 2 cm / 4 h = 0.5 cm/h
3. Determine the ERH by subtracting the infiltration (e.g., ø-index) from the GRH:
Module 3
Unit hydrograph
Example Problem Contd…
Time (hr) Effective precipitation (ERH)
(cm/hr)0-1 0
1-2 2
2-3 2
3-4 0
As observed in the table, the duration of the effective rainfall hyetograph is 2 hours.Thus, D = 2 hours, and the Unit Hydrograph obtained above is a 2-hour UnitHydrograph.
Module 3
Unit hydrograph
Example Problem Contd…
Lecture 5: Commonly used distributions in hydrology
Module 7
Normal Distribution
µσσ
α α
α µ α
σ
− −= ∏
≤ ≤
≤ ≤
>
∗
∗
21 1( ) exp 22
0
xf x
Also, called Saminion Distributon; Bell shaped Distribution
Most popular dis
- x +
- +
Normal Distribut
ion:
µ σ µ σ µ σ∗
2 2 2( , ), ( , )N N
tribution in any field
has two-parameters & denoted as X
x= μ
x
f(x)
+α-α
µσ−
= tan varxz where z is s dard normal iate Module 7
Normal Distribution Contd….
µ
α
κ
µ σ
→ ± →
=
→ 2
,
0
)
s
1) Symmetirc about x =
2) As x f(x) 0 asymptotically
3) C , as symmetric distrbution
4) = 3
5) If x Normal N( , and y = ax + b linear function
Properties of Normal Distribution:
µ σ+ 2 2 )b then y follows a normal dist. with N(a , a
Module 7
Central Limit Theorem
µσ
= + + + +
==
1 1 3 n
i i i2
i i
Here, Sn X X X X
n may not be very large, may be 6 or 7
For most hydrological applications, under some general conditions, it is shown that if X is an independent with E(X )and var(X ) ,
µ σ= =
= + + + +
∑ ∑1 1 3 n
n n2
i ii 1 i 1
then,
Sn X X X X
approches a normal distribution with E(Sn) = and var(Sn) =
i2i
2
If Sn is the sum of n i.i.d random variables Xi each having a mean µ and variance then, Sn approaches a normal distribution with mean, n and variance n as napproaches to infinity.
σµ σ
Module 7
Uniform Distribution
( )
α α α
β
α
α ββ α
β α β α
α α α ββ α β α β α
= ≤ ≤−
= = = − −
−− = ≤ ≤ − − −
∫ ∫
∫
1( )
1( ) ( )
( )
xx x
f x
xf x f X dx dx
x x
xf X dx
x
Probability is uniformly distributed
= for x
Expected value of x = E(x) =
β
α
β αβ α β α β α
β α β α β αβ α
= − − − −
+ − +−
2 2 2
2( ) 2( ) 2( )
( )( ) ( )2( ) 2
x =
= =
Module 7
Uniform Distribution Contd….
[ ]
( )( )
( )( )
( )( )
( )
( )
22 2
3 322
3 3 22
3 3 2 2
3 3 2 2 2 3 3 2 2
3 3 2 2
( ) ( )
( )3
,( )
3 2
4 3( 2 )( )
124 4 3 6 6 3 3 6 3
123 3
12
E x E x
xE x dx
Now
β
α
σ
β α
β α β α
β α β ασβ α
β α α αβ β β α
β αβ α α β αβ αβ β α α β β α
β αα β α β αβ
β α
∴ = −
− ∴ = =
− −
− + = − −
− − + + −=
−− − − − − + + +
=−
− + + −=
−
∫
Module 7
Uniform Distribution Contd….
( )( )
( )( )( )
( )( )
( ) ( )
3 3
2 2
2 2
2 2
2 22
3 ( )
123 ( )
123
122
12
12 12s
α β αβ α β
β αα β α αβ β αβ α β
β αα β α αβ β αβ
β αα αβ β
α β α β
− + + −=
−− − + + + −
=−
− − − − + =−
− − + =
− − −= = =
−
Module 7
Exponential Distribution
λ
λ
λ
α αλ
αλ
αλ λ
λ λ
λ
λ
λ
λ
λλ λ
−
−
−
−
−
− −
=
= =
= −
= =
=
= − − −
∫ ∫
∫ ∫
∫
∫
0 0
0 0
0
0
( ) x>0; >0
If you set ( ) ( )
1 x>0 & >0
( ) ( )
1.
x
x xx
x
x
x
x x
f x e
F x f x dx e dx
e
E x xf x dx x e dx
xe dx
xe e dx
αλλ
α
λ λ
λ
λ
λ
λλ
λ
−−
+
= − −
= − +
= − + − −
= =
∴ =
0
0
2
1
1 0 0 0
1 1 finally & ;
1 variance
xx
x x
exe
xe e
x
Module 7
Gamma Distribution
( )
η η λ
α
λ λ η η ηη
η η η
η η η η
η η
− −
− −
= Γ =Γ
Γ = −
Γ + = Γ
Γ =
Γ = Γ = Γ = Π
∫
1
1
0
( ) ; x, , >0 ( ) Gamma function of ( )
1) ( ) 1 ! f
Propertie
or =1,2, ....
2) ( 1) ( ) for >0
3) ( ) for >0
14) (1) (2) 1 )2
:
; (
s
x
n t
x ef x
t e dt
Module 7
Gamma Distribution Contd….
When data are positively skewed we have to use Gamma distribution.
η =0.5λ = 1
η = 1λ = 1
η = 3λ = 4
η = Scale parameterλ = Shape parameter
η γλ η
ησλ
= = =
=22
2( ) ;
x
E x Coefficient of Skewness
Module 7
Log-normal Distribution
x
= +
= + =
2
2
2 2
We say x is log-normally distributed, if y = ln X is normally
1 ( 1)
ln(1 )
d
istributed.
v
xy v
xy Cv is Coefficient of VariationCz
SS C Cvx
µα
σσ
µσσ
=
= + + + −− ∴ = = = ≤ ≤ ∏
−− = ∴ = = ∏
1 2 n
1 2 n2
y y
yy2
y
yy
If x x x x or
ln(x) ln(x ) ln(x ) ln(x )y1 1f(y) exp y ln(x) or x e [0 y ]22
ln(x)dy 1 dy 1 1Now, f(x) f
Here
(y) exp
, post
2dx x
ive
dx 2 .x.
skewn = + 2s v v sess r 3C C , as Cv increses, r also increses.
Module 7
Extreme Value Distribution
Distribution depends on
Number of random variables considering the point set of distribution
Parent distribution of the RVs (X1, X2, ……Xn)
[ ]
[ ]
[ ]
= ≤
= ≤
= ≤ = ≤
≤
If y is the max. value (if it also a RV)
( )
We know ( ) , f(y) corresponding pdf
( ) [ ]
(becomes max. value y)
i
y
x i
y i
F y P Y y
F x P X x
F y P Y y P All X y
Module 7
Extreme Value Distribution Contd…
[ ] [ ] [ ] [ ]
[ ]
→
∴ = ≤ × ≤ × ≤ × ≤
= × × × == =
1 2
1 2
1 2
1 2 3
(i.i.d independent and identically distributed)
Let X ,X ,..................X be i.i.d
( ) .........
( ) ( ) ........ ( ) ( ) ( ) ( ) ..
n
n
y n
nX X X X
X X
F y P X y P X y P X y P X y
F y F y F y F yF y F y = ..... ( )
nXF y
[ ] [ ]
[ ]
− −
∂ ∂ = ∂ ∂
∂ = = ∂
∴ =
1 1
( )( ) is pdf of y = ( )
n ( ) ( ) n ( ) ( )
( ) ( ) ( )
nyy y
n nX y X x
ny x x
F yf y F y
y y
F y F y F y f yy
f y n F y f y
Module 7
Lecture 6: Goodness of fit tests
Module 7
The goodness of fit tests measure the compatibility of a random sample
with a theoretical probability distribution function. In other words, these tests
show how well the distribution one selected fits to the data.
Kolmogorov-Smirnov Test
This test is used to decide if a sample comes from a hypothesized
continuous distribution. It is based on the empirical cumulative distribution
function (ECDF). Assume that we have a random sample x1, ... , xn from
some distribution with CDF F(x). The empirical CDF is denoted by
( )nF ( x ) Number of observations xn
= ≤1
Goodness of Fit
Module 7
The Kolmogorov-Smirnov statistic (K) is based on the largest vertical difference
between the theoretical and the empirical cumulative distribution function:
Hypothesis Testing
The null and the alternative hypotheses are:
H0: the data follow the specified distribution;
HA: the data do not follow the specified distribution.
The hypothesis regarding the distributional form is rejected at the chosen
significance level (α) if the test statistic, K, is greater than the critical value. The
fixed values of (0.01, 0.05 etc.) are generally used to evaluate the null hypothesis
(H0) at various significance levels. A value of 0.05 is typically used for most
applications.
( ) ( )i ii n
i iK F x , F xn nmax
≤ ≤
− = − − 1
1
Module 7
Anderson-Darling Test
The Anderson-Darling procedure is a general test to compare the fit of
an observed cumulative distribution function to an expected cumulative
distribution function. This test gives more weight to the tails than the
Kolmogorov-Smirnov test.
The Anderson-Darling statistic (A2) is defined as:
( ) ( )( )( )n
i n ii
A n i . ln F( x ) ln F xn − +
=
= − − − + − ∑21
1
1 2 1 1
Goodness of Fit Contd…
Module 7
Chi-Squared Test
The Chi-Squared test is used to determine if a sample comes from a
population with a specific distribution. This test is applied to binned data, so
the value of the test statistic depends on how the data is binned. Please note
that this test is available for continuous sample data only.
The Chi-Squared statistic is defined as
where Oi is the observed frequency for bin i, and Ei is the expected frequency for
bin i calculated by, Ei =F(x2)-F(x1), where F is the CDF of the probability
distribution being tested and x1 and x2 are the limits for bin i.
( )ni i
ii
O E,
Eχ
=
−= ∑
2
2
1
Goodness of Fit Contd…
Module 7
The relative goodness of a model fit may be checked using
RMSE (Root Mean Square Error),
AIC (Akaike Information Criterion) and
BIC (Bayesian Information Criterion).
RMSE
The root mean square error is defined as the square root of mean sum of
square of difference between empirical distribution and theoretical
distribution, i.e. MSERMSE =
Module 7
Goodness of Fit Contd…
AIC
The AIC is used to identify the most appropriate probability distribution.
It includes:
(1) the lack of fit of the model and
(2) the unreliability of the model due to the number of model parameters
It can be expressed as:
AIC = -2(log(maximum likelihood for model)) + 2(no of fitted parameters) or
AIC = N log(MSE) + 2(no of fitted parameters)
Goodness of Fit Contd…
Module 7
BIC
The BIC (Schwarz, 1978) is another measurement of model selection which
can be expressed as :
BIC = N log(MSE) + [(no of fitted parameters)* log(N)
The best model is the one which has the minimum RMSE, AIC and BIC values
Goodness of Fit Contd…
Module 7
Lecture 7: Statistical parameter estimation
Module 7
Parameter Estimation
Methods of Parameter Estimation
1) Method of Matching Points
2) Method of Moments
3) Maximum Likelihood method
θ
θ
θθ θ
→
→
=
Population Parameter
Sample Parameter
Unbiased estimation of parameter:An estimate of a parameter is said to be unbiased estimate, if E( )
i
i
i
i i
Module 7
1) Method of Matching Points
θ
θ
θθ
θ
−
−
∴ ≤ ∫ ∫0 0
In a data set, 75 % values are less than, It is assumed to follow the distribution,
f(x) = ; x > 0, estimate the parameter .
P[X 3] = 0.75 = F(x)= ( ) =
x
xx x
e
ef x dx dx
θθ
θ
θ θ
θθ
−−
−
− ×
−∴
0
= = 1-e1( )
1 - e = 0.75 or = -1.3863 or = 2.164
xx
x
x
e
x
Module 7
2) Method of Moments
θ θ θ θi j i j
1 2 n
Given a function f( ,......., ,x) and values ,.......,
we need to find x ,x ,..........x Generate number of equations by taking moments of the distribution
Take, any distribution, like f(xθ
θθ α α− −
−= ∏
⇒
212
2
( x )1 22 2
2) (2 ) e - < x < +
Take the 1st moment Mean about the originθα
θ
α
θαθ
α
µ θ
θ
− −−
−
− −−
−
= ∏
∏
∫ ∫
∫
212
2
212
2
( )1 22 2
2
( )1 22 2
2
( ) (2 )
(2 )
x
x
x fx dx x e dx
xe dx
E(X) = =
=
Module 7
2) Method of Moments Contd...
α
α
θ θ θθ
−
−
+∏ ∫
2
22 1 2
2
1 ( )2
y
y e dx E(X) =
θθ
θ θ
θ
−
+
1
2
2 1
2
( )substituting, y = ( )
x =
dx =
x
y
dy
α α
α α
α
α
θ θ
θ
− −
− −
−
−
+
∏
⇓
∏
∫ ∫
∫
2 2
2
2 22 1
21
12
1 02
y y
y
e ydy e dy
e dy
E(X) =
E(X) = +
θθ
θ µ
∏∏
∴ = =
11
1
22
( )E x
[As odd multiplier, h(-y) = -h(y)
= 0 + =
Module 7
2) Method of Moments Contd…
Second moment about the mean,
E (X- ) =
=
y = and we will get,
212
2
2 2 2
( )1 22 2 2
2
1
2
( ) ( )
( ) (2 )
,
2
x
x f x dx
x e dx
Substituting
xand
α
α
θαθ
α
µ σ µ
µ θ
θ µ
µθ
−
− −−
−
= −
− ∏
=
−
∫
∫
22σ θ=
Module 7
3) Maximum Likelihood method
θ
θ θ θ
θ
=
1 1
1 2 3
,
We have the following, ( ; ) ( ; ) ( ; )
i i
Samplex
f x f x f x
x
1 3Product of ( ; ) ( ; ) is "likelihood " L
If L( ; ) ( ; ), then is the estimate preferred,
which maxmizes the likelihood func
tion.
θ θ
θ θ θ
× × ≈
>
∗
∗
i i
f x f x
x f x
θ →( ; ) evaluated at x = x i if x pdf
Module 7
3) Maximum Likelihood method Contd…
{ }
1 2
1 2 ) 1
1 2
1 2 3
(
( ) ; 0 is a parameter
, , , Sample available;
L = ( , ) ( , ) ( , ) ( , )
=
= = (formulation of like
n
n
in i
x
n
n
x x x
xx x xn n
f x e x
x x x
f x f x f x f x
e e e
e e
β
β β β
ββ
β β
β β β β
β β β
β β =
−
− − −
−− + + +
= >
→
∴ × × × ×
× × ×
∑
lihood function)
Because ln(L) is an increasing function of L, it reaches maximum value
ln( )at the same pt., as ln(L) does, 0
(When there is no other method feasible, this method is be
st one)
θ∂
=∂
⇒
∗
L
Module 7
3) Maximum Likelihood method Contd…
Now set L maximum
1
1
1
ln( ) ln
ln( ) 0 as we want to set the value of
ln( ) 0
1 = Arithmetic average
Max. likelihood estimate
n
ii
x
n
ii
n
ii
L n e
L
L n x
n xxx
β β
ββ
β β
β
=
=
=
∑∴ = −
∂∴ =
∂
∂∴ = − =
∂
∴ = =
∑
∑
Module 7
3) Maximum Likelihood method Contd…
( )
µσσ
σ µ
µµ σ µ σ µ σ
σσ
µσ
σ=
− −= ∏
− −∴ ∏
− −= − ∏ −
2
2
1 2
22
1 1 ( ) exp 22 [Take, as parameter not S.D. and also]
1 1L = ( , , ) ( , , ) ( , , ) = exp 22
1 ln( ) ln(2 ) 22
in n
i
i
xf x
xf x f x f x
xnor L
µ σµ σ
µµ
µ σ=
∂ ∂= =
∂ ∂
−∂ = = ∴ − = ∂
∑
∑ ∑
1
21
ln( ) ln( ) 0 set & ,Now
ln( ) 0 ( ) 0
n
ni
ii
L Lor
xL X
= 1
n
ii
xor x
nµ = =
∑
Module 7
3) Maximum Likelihood method Contd…
( )
2
2
1 2
22
1 1( ) exp 22
1 1( , , ) ( , , ) ( , , ) exp 22
1ln( ) ln(2 ) 22
µσσ
σ µ
µµ σ µ σ µ σ
σσ
µσ
σ
− −= ∏
− −∴ ∏
− −= − ∏ −
[Take, as parameter not S.D. and also]
L = =
in n
i
xf x
xf x f x f x
xnor L1
21
ln( ) ln( )0
ln( ) 0
( ) 0
µ σµ σ
µµ σ
µ
=
=
∂ ∂= =
∂ ∂
−∂ = = ∂
∴ − =
∑
∑
∑
set & ,Now
n
i
ni
i
i
L Lor
xL
X
Module 7
3) Maximum Likelihood method Contd…
µ
µσµ σ σ
µσ σ
µσ
=
=
=
=
=
− −∂ ∏ = = − − ∂ ∏
−+ =
−=
∑
∑
∑
∑
1
2
2 31
2
31
2
2 1
( ) ( 2)ln( ) 4 102 22
( )0
( )
n
ii
ni
i
ni
i
n
ii
xor x
n
xL nand
xnor
xor
n
=
-
[But it is not the best method. It depends upon situation]
Module 7
In hydrology, most of the phenomena are random in nature.E.g. rainfall-runoff model
Random variables involved in a hydrological process may be dependent or
independent.
The ‘random variables’ X & Y are ‘stochastically independent’ if and only if
their ‘joint density’ is equal to the product of ‘marginal density functions’.
Joint density function : Simultaneous occurrence
Marginal density function : Distribution of one variable irrespective of the value
of the other variables
Conditioned distribution: Distribution of one variable conditioned on the other
variable.
Highlights in the Module
Module 7
Measures of Central Tendency:
Mean
Arithmetic average (for sample)
Mode
Median
Measures of Spread or Dispersion:
Range [(xmax-xmin)]
Relative Range [=(range/mean)]
Variance
Highlights in the Module Contd…
Module 7
Standard deviation,
Coefficient of variation
Highlights in the Module Contd…
Measures of Symmetry:
Coefficient of skewness,
Kurtosis
Correlation coefficient shows the degree of linear association between
two random variables
Commonly used distributions in hydrology :
Normal distribution
Uniform distribution
Exponential distribution
Gamma distribution
Log-normal distribution
Extreme value distribution
Module 7
Methods of statistical parameter estimation
1) Method of Matching Points
2) Method of Moments
3) Maximum Likelihood method
Highlights in the Module Contd…
Module 7
Hydrological Statistics(contd.)
Prof. Subhankar KarmakarIIT Bombay
Module 85 Lectures
Objective of this module is to learn the concepts instochastic hydrology.
Module 8
Topics to be covered
Frequency analysis
Markov process
Markov chain
Reliability analysis
Module 8
Lecture 1: Frequency analysis
Module 8
Frequency Analysis
[ ]
[ ]
1 1
1
= + = +
= +
v
T T v
We can express X = X+k S, where k can bepositiveor negativeno and S is SD.k S X X kCX
In general, X X k C and T is any
Return Period [T
giv
]:
en time
Expected i
..
1
nterval between successive occurences of an event (in a long interval)
p= where p is probability of occurence of the event in any year and T
T is in years,
[ ] 1≥ = P X
Where is magnitude of the T year event (flood)
T T
T
xT
xModule 8
Frequency Analysis Contd…
[ ]1= +
=
=
=
Our problem is, given T, how to find X
From , X
K function of distribution and return period
from data
K Frequency factors [Function of distribution and & return
T
T T v
T
v
T
?
X k C
C
[ ]1= +
+−
= =
period, T]
Normal dist. X
=
(Std. normal variate)
T T v
T
TT
X k C
X SkX Xor k Z
S
Module 8
Treatment of Zeros
[ ] [ ]1
0 0
=
×
= ≠
∑
(B1, ....., B6) &
Let A be an event within that region.
stat
Mutually exclusive events collectively exhaustive
Total Probability th es:
P[A] =
are mutually exclusive and co
eorem
lle
n
i ii
P A | B P B
[ x & x
≠
ctively exhaustive]
Define, x=0 & x 0 as two mutually exclusive and collectively exhausted events
B2 A B3
B4
B5B6
B1
0
P[x=0]f(x|x≠0)
Treatment of Zeros Contd…
[ ] [ ] [ ] [ ] [ ]
[ ] [ ]
[ ]
[ ]
[ ]
≥ = ≥ = × = + ≥ ≠ × ≠
= ≥ ≠ × ≠
= ≥ ⇒
= ≤ ≠ ⇒
= ≠
− = − ×
Using this notation
T
*
*
P X x P X x | X P X P X x | X P X
P X x | X P XNotations :
i )F( x ) P X x without conditions CDF
ii )F ( x ) P X x | X CDF conditional
iii )K P X
F( x ) F ( x ) k
or F( x
0 0 0 0
0 0
0
0
1 1
= − +
where, F(x) includes all the zeros in the data (non conditional)
and F*(x) does not include zeros (only non-zero data)
*) k kF ( x )1
Module 8
Example Problem 1
If in a sample there are 95% non-zero values, calculate X .10
[ ]
[ ]
σ
= ≤ ≠ ≠
= ≤
∴
−= = × + =
*
*
*
F ( x ) . P X x | X
If F ( x )
F ( x ) . P Z z
xor . or . .
10
2
0 895 0
0 895
1255 1255 15 10 28 33Tt
= given x 0
follows a normal distribution N (10,15 ) given
=
Get z value corresponding to 0.895 z = 1.255
x x units
[ ] [ ]Solution:= = ≠ ≥ = =
= − = ⇒ = − + *
k . P X , P X x .
F( x ) . . . . . F ( x )10
10 10
10 95 0 0 110
1 0 1 0 9 0 9 1 0 9 5 0 9 5
Module 8
Peak flow data are available for 75 yrs, 20 of the values are zero and the
remaining 55 values have a mean of 100 units and std. deviation of 35.1 units
and are log normally distributed. Estimate the probability of the peak
exceeding 125 units using frequency analysis.
Example Problem 2
[ ]
[ ]
( )
55 073375
125 1 1 125 1
125
125 0
1
= = ≠
> = = − = − +
=
≤ ≠
= +
T
T
P[X 0]
log-normally distributed
=
table, frequency table
For normal dist. it is K = S
X For l
*
* *T
T
T V
k .
P X F( ) F( x ) k kF ( x )
F ( X ) F ( )
P X X
K
X K C og-normal dist.
Module 8
[ ]
125 100 1 0 351
351 100 0 351
0 712
0 21 0 79
1 0 733 0 733 0 790 846
1 0 846 0154
V
C
But we are interested in F(x)
= +
= = = =
=
∴ ≥ = =
= − + ×=
≥ = − =
T
T
T
T
or ( . K )
SCoV . / .X
or K .
P[ X X ] . .
. . ..
P X X . . ( Ans )
Module 8
Example Problem 2 Contd…
Lecture 2: Markov process and Thomas Fierring model
Module 8
First Order, Stationary Markov Process
1 1 1( )t x t x iX Xµ ρ µ+ += + − +∈
2
Assuming Dependence to random component with Stationary previous observation zero mean and variance
= onlyx
σµ µ
∈
1Estimate of X
nt
i
xn µ
=
= =∑Module 8
Many hydrologic time series exhibit significant serial correlation.
Value of the random variable under consideration, X at one time period (say, t+1)
is correlated with the values of the random variable at earlier time periods.
[ ]{ } ( ) [ ]{ } ( )
( ) [ ] ( ) [ ]{ } ( )
( )
2 22 221 1 1 1 1
22 221 1 1 1 1
2 2 2 2 2 2 2 21 1
2 2 21
, ( )
2
= 0 ( )
, 1 ...
x t t x t x i t
t x x i t x x i t
x x x x
x
Now E X E X E x E X
E x x E X
So
σ µ ρ µ
ρ µ µ ρ µ µ
ρ σ µ σ µ ρ σ σ
σ σ ρ
+ + + +
+ + +
∈ ∈
∈
= − = + − +∈ −
= − + +∈ − + +∈ −
+ + + − = +
= −
[ ] ( )1 1 1
1 1
...random component for first order stationary markov process
[Check:
0 so first order stationary]
X N(
t x t x i
x x x
x
E X E x
if
µ ρ µ
µ ρ µ µ ρµ
+ + = + − +∈ = + − +
=
2
2
, ), (0, )
x x
then Nµ σ
σ∈∈
First Order, Stationary Markov Process Contd…
Module 8
( )( )
1
N
2N
21
21 1 1
1) ,
2) , R (0,1)
0such that R (0, )
= 1
( ) 1
3) If negative value comes, retain it for gent t t
x x
N N x
t x t x N x
Estimate and
Generate random numbers from the distribution N
as N
or R R
X x R
µ σ ρ
σσ
σ σ ρ
µ ρ µ σ ρ
∈
∈
∈
+
∈ −= ∈
∈ = −
∴ = + − + −
erating the next value, but discard it in application.4) Generate a large no. of values and discard the first 50-100 values to ensure that there is no effect of initial value.
First Order, Stationary Markov Process Contd…
Module 8
To generate values for Xt+1
First Order, Stationary Markov Process Contd…
For log-normally distributed data
Preserves the mean, std. deviation and lag 1 correlation of logarithms of
flows (not the original data)
This method may be applied on annual runoff, annual rainfall etc. (but not
on 10 day rainfall series)
{ }
( )21 1 1
ln
( ) 1
t t
t y t y N y y
Y X
Y y Rµ ρ µ σ ρ+
=
= + − + −
Module 8
First Order Markov Process(Thomas Fierring Model)
Consider monthly stream flow data for n years
m: no. of values (months)
Here, i= 1,2…, n and j= 1,2,…,m.
( )( ) ( )1 2
, 1 1 , 1 1
: Serial correlation between jth month and (j+1)th month
when j+1 = (12+1)=13 implies that i=2, j=1
1
j
ji j j j ij j i j j j
j
X X t
ρ
σµ ρ µ σ ρ
σ+
+ + + +
∴ = + − + −
Module 8
Month SD rj1 15.7 4.14 0.8642 13.62 4.15 0.3023 26.21 24.32 0.854
12 18.01 5.08 0.637
X
11 1One assumption, X X Xµ= =
correlation with next month
This preserves seasonal mean, standard deviation and lag1 correlation.
Non-stationarity in mean and std. deviationModule 8
Thomas Fierring Model
Lecture 3: Markov chain
Module 8
• A Markov chain is a stochastic process having the property that the value of
the process Xt at time t, depends only on its value at time t-1, Xt-1 and not on
the sequence Xt-2, Xt-3, ……, X0 that the process passed through to arrive at
Xt-1
For 1st order Markov chain or Single step M.C.
Module 8
( )( )
1 2 3 1 0
1
prob | , , ,...,
prob |t j t i t k t q
t j t i
X a X a X a X a X a
X a X a− − −
−
= = = = =
= = =
Markov Chains
Markov Chains Contd….
Xo Xt-2 Xt-1 Xt
Module 8
Diagrammatically, it may be represented as,
We will be able to write this as
Xt-1=ai
t-1
Xt=aj
t time period
State i transited to State j
1
is the probability that it goes in to state j, starting with array i here.
tij t j t ip P X a X a
P− = = =
Markov Chains Contd….
Transition probability
It is the probability that state ‘i’ will transit to state ‘j’
[i.e., transition probabilities remain the same across the time]
t-3 t-2 t-1
pt-2ij
t
ptij
then, the series is called homogeneous Markov Chain t,t tij ijIf p p τ τ+= ∀
Module 8
Here analysis is done only for:
Single step (1st order) homogeneous M. C.
If t is month then pij will not be homogeneous (seasonal change)
pij = transition probability for ‘i’ to ‘j’
i=1,2, ……, m and j=1,2,…….,m
where m is the no. of possible states that the process can occupy.
Markov Chains Contd….
Module 8
11 12 13 1m
21 22 23 2mij
m1 m2 m3 mm
TPM, (Total probability matrix)p p p pp p p p
P = p =
p p p p
Probability stating that i=1 go into j=2
Sum of each row=1
Markov Chains Contd….
m
j=1
Each row must add to 1
t,
1,
t tij ij
ij
If p p
p i
τ τ+= ∀
∴ = ∀∑
Such matrices whose individual rows add up to 1 are called the “stochastic matrices”
Module 8
2
1
( 1), probability values that need to be estimated
Estimate from historical data
total no. of
ij
ijmij
ijj
m m m m
p
np
n
∧
∧
=
− = −
=∑
Markov Chains Contd….
Historic data
1 31 2 1 2
100Time period
1 2
States (random)e.g. No. of times it went into state 1 out of these 50 times = 20,
No. of times it transited to state 2 = 20
No. of times it transited to state 3 = 15 Then
12
15 ;50
p∧ =
Deficit, non-deficit Two statesDrought, non-drought Two states
Module 8
11
20 ; 50
p∧ =
13
15 .50
p∧ =
Markov Chains Contd….
• p(n)j : probability that the process will be in state j after n time steps
i j - state
nTime interval
t-1
• p(0)j : Initial probability of being in state j
( ) ( ) ( )1 2 1
a sum vector, ,............, .....n n n nm xm
p p p p=
Probability of being in state 1 in time step n
Module 8
Markov Chains Contd….
(1)
(1) ( )
11 12 13 1m
21 22 23 2m(0) (0) ( )1 2
m1 m2 m3 mm
is given create at t=1
from probabilityp p p pp p p p
= , ,...,
p p p p
→
=
×
o
o
om
P
If pp
p p
p p pProbability vector at time 1
+ + + (0) (0) (0)1 11 2 21 1............ m mp p p p p p
Probability that event will start from state ‘2’
Probability of transition from 2 to 1
Module 8
Markov Chains Contd….
Probability that the state is 1 in period 1
= (1) (1) (1)1 2, ,............, mp p p
Probability that state is ‘2 ‘ in time period ‘1’
Probability that state is ‘m in time period ‘1’
(2) (1)
(0)
(0) 2
, .
= . . =
Simillarlyp p P
p P Pp P
=( ) ( 0 )
Any time period n,
n np p P=
After time n, prob. to be in particular state ‘j’
Initial prob. vector
(TPM)
Module 8
Markov Chains Contd….
(n m) (n)
(n m) (n)
p p ,after a large m then steady state probabilitycondition is achieved.Once the steady state is reached,p p pSo,p p.P
+
+
=
= =
=
Module 8
Example:
• A 2-state Markov Chain; for a sequence of wet and dry spellsi = 1 dry;i = 2 wet
0.9 0.10.5 0.5
P =
d
w
wd
(i) P [day 1 is wet |day 0 is dry]
= P [Xt = 2|Xt-1 = 1]
= p12 = p(1)2 = 0.1 (Ans)
Example Problem
(ii) P[day 2 is wet |day 0 is dry]
[ ]
[ ]
=
=
(2)2
(2) (1)
(2)
= 0.9 0.1
0.86 0.14
pp p P
p
day
wetBecause day 0 is dry
Dry wet
Probability that day 2 will be wet
= =(2)2 0.14p
Module 8
0.9 0.10.5 0.5
(iii) Prob [day 100 is wet |day 1 is dry]
Example Problem Contd….
(100)2. .,0.9 0.10.5 0.5
i e P
P =
Module 8
Here, the fact that day 1 was dry, would not significantly affect the probability of
rain on day 100. So n can be assumed to be large and solve the problem based
on steady-state probabilities
2
4 2 2
8 4 4
16 8 8
To determine steady state,0.86 0.14
or P . .0.7 0.30.8376 0.1624
P . .0.8120 0.18800.8334 0.1666
P . .0.8320 0.16720.8333 0.1667
P . .0.8333 0.1667
P P
or P P
or P P
or P P
= =
= =
= =
= =
∴All the rows same
p=(0.8333; 0.1667)
dry wet
Module 8
Example Problem Contd….
Lecture 4: Data generation
Module 8
Data Generation
Module 8
Data generation from a Markov chain requires only a knowledge of the initial
state and the transitional probability matrix ,
To determine the state at time 2, a random number is selected between 0
and 1.
P
NR from N(0,1).
n-1 n-1
N ij ij j=1 j=1
If this R is between p and p , the state is 'n'∑ ∑
Module 8
Consider a Markov chain model for the amount of water in storage in a reservoir.
Let state 1 represent the nearly full condition, state 2 an intermediate condition and
state 3 the nearly empty condition. Assume that the transition probability matrix
is given by,
Note that it is not possible to pass directly from state 1 to state 3 or from state 3 to
state 1 without going through state 2. Over the long run, what fraction of the
time is the reservoir level in each of the states?
1 2 31 0.4 0.6 02 0.2 0.6 0.23 0 0.7 0.3
P =
Example Problem 1
Module 8
Example Problem 1 Contd…
Solution:
Let p1= 1 p2 = 3 and p3 = 6/7 . After scaling, the solution can be written as:
(p1, p2, p3) = (.2059, .6176, .1765).
Thus, over the long run, the reservoir is nearly full 20.59% of the time, nearly empty 17.65% of the time and in the intermediate state 61.76% of the time.
( ) ( )1 2 3 1 2 3
1 2 1
1 2 3 2
2 3 3
p.P p
.4 .6 0p p p .2 .6 .2 p p p
0 .7 .3.4p .2p 0 p.6p .6p .7p p0 .2p .3p p
=
=
+ + =
+ + =
+ + =
Module 8
Example Problem 2
Assume that the reservoir of previous example is nearly full at t=0. Generate a sequence of 10 possible reservoir levels corresponding to t=1, 2,…, 10.
Solution:
At t=0, state is 1.Generate cumulative transition probability matrix (CTPM) where
* * *
1
*
,0.4 1.0 00.2 0.8 1.00 0.7 1.0
k
ik ik ijj
P P andP p
Here
P
=
= =
=
∑
*P
t State at t RN State at t+1 Reservoir level at t
0 1 0.48 2 nearly full1 2 0.52 2 intermediate2 2 0.74 2 Intermediate3 2 0.15 1 intermediate4 1 0.27 1 nearly full5 1 0.03 1 nearly full6 1 0.49 2 nearly full7 2 0.02 1 intermediate8 1 0.97 2 nearly full9 2 0.96 3 intermediate
10 3 nearly empty
Module 8
Example Problem 2 Contd…
Lecture 5: Reliability analysis
Module 8
Reliability
It is defined as the probability of non-failure, ps, at which the resistance of the
system exceeds the load;
where P() denotes the probability.
The failure probability, pf , is the compliment of the reliability which can be
expressed as
)( RLPps ≤=
sf pRLPp −=≥= 1)(
The resistance or strength (R) is the ability to accomplish the intended mission
satisfactorily without failure when subjected to loading of demands or external
stresses (L). Failure occurs when the resistance of the system is exceeded by the
load (floods, storms etc.)Module 8
Measurements of Reliability
Recurrence interval
T=1/(1-F), F= P(X<xT) No consideration for the interaction with the system resistance
Safety Margin
It is defined as the difference between the resistance (R) and the
anticipated load (L)
SM=R-L Safety Factor
It is the ratio of resistance to load
SF=R/L(Tung, 2004)
Module 8
Recurrence Interval
Assume independence of occurrence of events and the hydraulic structure design for an event of T-year return period.
1/T is the probability of exceedance for the hydrologic event in any one year. Failure probability over an n-year service period, pf, is
pf = 1-(1-1/T)n (using Binomial distribution)or pf = 1-exp(-n/T) (using Poisson distribution)
Types of problem:(a) Given T, n, find pf
(b) Specify pf & T, find n(c) Specify pf & n, find T
Module 8
Probabilistic Approaches to Reliability
Statistical analysis of data of past failure records for similar systems
Reliability analysis, which considers and combines the contribution of
each factor potentially influencing the failure with the steps as
(1) to identify and analyze the uncertainties of each contributing factor;
and
(2) to combine the uncertainties of the stochastic factors to determine
the overall reliability of the structure.
Module 8
Uncertainties in Hydraulic Engineering Design
Hydrologic uncertainty (Inherent, parameter, or model uncertainties)
Hydraulic uncertainty (Uncertainty in the design and analysis of hydraulic structures)
Structural uncertainty(Failure from structural weaknesses)
Economic uncertainty(Uncertainties in various cost items, inflation, project
life, and other intangible factors)
Techniques for Uncertainty Analysis
Analytical Technique
Fourier and Exponential Transforms
Mellin Transform
Approximate Technique
First-Order Variance Estimation (FOVE) Method
Rosenblueth’s Probabilistic Point Estimation (PE) Method
Harr’s Probabilistic Point Estimation (PE) Method
Reliability Analysis Methods
Module 8
Reliability Analysis Methods
1. Performance Function and Reliability Index
2. Direct Integration Method
3. Mean-Value First-Order Second-Moment (MFOSM) Method
4. Advanced First-Order Second-Moment (AFOSM) Method
a) First-order approximation of performance function at design point.
b) Algorithms of AFOSM for independent normal parameters.
c) Treatment of correlated normal random variables.
d) Treatment of non-normal random variables.
e) AFOSM reliability analysis for non-normal, correlated random variables.
5. Monte Carlo Simulation Methods
1. Performance Function and Reliability Index
To enable a quantitative analysis of the reliability of a structure, every failure
mode has to be cast in a mathematical form.
A limit state function is given by:
W (z, x) = R (z, x) – S (z, x)
where,
z: Vector of design variables;
x: Vector of random input variables;
R: Resistance of the structure;
S: Load on the structure.
The value of the limit state function for given values of x and z denotes the
margin.
The reliability index is defined as the ratio of the mean to the standard
deviation of the performance function W(z, x)
where μW and σW are the mean and standard deviation of the
performance function.
The boundary that separates the safe set and failure set is the failure
surface, defined by the function W(z, x) =0, called the limit state
function.
ww σµβ /=
1. Performance Function and Reliability Index Contd…
Module 8
System states defined by performance function
1. Performance Function and Reliability Index Contd…
Module 8
2. Direct Integration Method
The reliability can be computed in terms of the joint PDF of the load and resistance as
where fR,L(r, l ) : joint PDF of random load, L, and resistance, R;
r, l : dummy arguments for the resistance and load, respectively;
(r1, r2), (l1, l2) : lower and upper bounds for the resistance and load, respectively.
( ) ( ) dldrlrfdrdllrfpl
l
r
lLR
r
r
r
lLRs ∫ ∫∫ ∫
=
=
2
1
22
1 1
,, ,,
Module 8
3. Mean-Value First-Order Second-Moment (MFOSM) Method
Here, the performance function W(z, x), defined on the basis of the load and
resistance functions, S(z, x) and R(z, x), are expanded in a Taylor series at a
selected reference point.
The second and higher order terms in the series expansion are truncated,
resulting in an approximation that requires the first two statistical moments of
the random variables.
The simplification of Taylor series greatly enhances the practicality of the first
order methods because in many situations, it is difficult to find the PDF of the
variables while it is relatively simple to estimate the first two statistical moments.
Module 8
4. Advanced First-Order Second-Moment (AFOSM) Method
Here, it mitigates the deficiencies associated with the MFOSM method, while
keeping the simplicity of the first–order approximation.
The difference between the AFOSM and MFOSM methods is that the
expansion point in the first–order Taylor series expansion in the AFOSM
method is located on the failure surface defined by the limit state equation,
W(x) =0.
Module 8
5. Monte Carlo Simulation Methods
It is a general purpose method to estimate the statistical properties of a
random variable that is related to a number of random variables which may or
may not be correlated.
The values of stochastic parameters are generated according to their
distributional properties and are used to compute the value of performance
function
Reliability of the structure can be estimated by computing the ratio of the
number of realizations with W ≥ 0 to the total number of simulated realizations.
Disadvantage computational intensiveness.
Module 8
Return period/recurrence interval: The reciprocal of annual exceedance
probability
First order, Stationary Markov Process states that observed value of
a random event at any time t+1, is a function of a stationary component,
correlated component and a purely random component. Here, the random
component is assumed to be normally distributed with zero mean and
variance.
2tσ
Highlights in the Module
Module 8
Reliability is defined as the probability of non-failure, ps, at which the
resistance of the system exceeds the load.
Failure occurs when the resistance of the system is exceeded by the load
(floods, storms etc.)
Measurements of reliability are recurrence interval, safety margin
and Safety factor
Techniques for Uncertainty Analysis:
Analytical technique
Approximate technique
Reliability analysis methods
Highlights in the Module Contd…
Module 8
Reliability Analysis Methods:
1. Performance Function and Reliability Index
2. Direct Integration Method
3. Mean-Value First-Order Second-Moment (MFOSM) Method
4. Advanced First-Order Second-Moment (AFOSM) Method
5. Monte-Carlo Simulation method
Highlights in the Module Contd…
Module 8
Module 95 Lectures
Hydrologic Simulation Models
Prof. Subhankar KarmakarIIT Bombay
Objectives of this module is to investigate on various hydrologic simulation models and the steps in watershed modeling along with applications and limitations of major
hydrologic models
Module 9
Topics to be covered
Hydrologic simulation models
Steps in watershed modeling
Major hydrologic models
HSPF(SWM)
HEC
MIKE
Module 9
Lecture 1: Introduction to hydrologic simulation modelling
Module 9
Watershed Classification
Watershed (ha) Classification
50,000-2,00,000
10,000-50,000
1,000-10,000
100-1,000
10-100
Watershed
Sub-watershed
Milli- watershed
Micro-watershed
Mini-watershed
(Bedient et al., 2008)
Module 9
A hydrologic simulation model is composed of three basic elements, which are:
(1) Equations that govern the hydrologic processes,
(2) Maps that define the study area and
(3) Database tables that numerically describe the study area and model
parameters.
Hydrologic Simulation Model
Equations Maps Database
Module 9
Hydrologic Simulation Model
A hydrological simulation model can also be defined here as a
mathematical model aimed at synthesizing a (continuous) record of some
hydrological variable Y, for a period T, from available concurrent records
of other variables X, Z, ... .
In contrast, a hydrological forecasting model is aimed at
synthesizing a record of a variable Y (or estimating some of its states) in
an interval ∆T, from available records of the same variable Y and/or other
variables X, Z, ... , in an immediately preceding period T.
Module 9
Hydrologic Simulation Model Contd…
A hydrological simulation model can operate in a "forecasting mode" if
estimates of the records of the independent variables (predictors) X, Z, ..., for the
forecast interval ∆T are available through an independent forecast. Then the
simulation model, by simulating a record of the dependent variable, [Y] ∆T will in
fact produce its forecast.
In short, a hydrological simulation model works in
a forecasting mode whenever it uses forecasted
rather than observed records of the independent
variables.
Module 9
Hydrologic Simulation Model Contd…
http://www.crwr.utexas.edu/gis/gishydro06/WaterQuality/GIStoHSPF/GIStoHSPF.htm
A Typical Watershed Delineation Model
Module 9
Hydrologic Model
Deterministic
Lumped
Steady Flow
Unsteady Flow
Event based Continuous
Distributed
Steady Flow
Unsteady Flow
Stochastic
Space Independent
Time Independent
Time Correlated
Space Dependent
Time Independent
Time Correlated
Classification of Hydrologic Models
Module 9
Spatial Scaling of Models
Module 9
LumpedParameters assigned to each sub-basin
Fully-DistributedParameters assigned to each grid cell
Semi-DistributedParameters assigned to each grid cell, but cells with same parameters are grouped
A1 A2
A3
1. Size
2. Shape
3. Physiography
4. Climate
5. Drainage
6. Land use
Parameters of Watershed
7. Vegetation
8. Geology and Soils
9. Hydrology
10. Hydrogeology
11. Socioeconomics
Module 9
Precipitation
Interception Storage
Surface Runoff
Groundwater Storage
Channel Processes
InterflowDirect Runoff
Surface Storage
BaseflowPercolation
Infiltration
ET
ET
ET: evapo-transpiration
Module 9
Flowchart of simple watershed model (McCuen, 1989)
Diversity of the current generation of models
There exists a multitude of watershed models, and their diversity is so large
that one can easily find more than one watershed model for addressing any
practical problem.
Comprehensive Nature
Many of the models can be applied to a range of problems.
Reasonable modeling of physical phenomena
In many cases models mimic reasonably well the physics of the underlying
hydrologic processes in space and time.
Strengths of Watershed Models
Module 9
Distributed in Space and Time
Many models are distributed in space and time.
Multi-disciplinary nature
Several of the models attempt to integrate with hydrology :a) Ecosystems and ecology,b) Environmental components,c) Biosystems,d) Geochemistry,e) Atmospheric sciences and f) Coastal processes
This reflects the increasing role of watershed models in tackling
environmental and ecosystems problems.
Module 9
Strengths of Watershed Models Contd…
The most ubiquitous deficiencies of the models are:
Lack of user-friendliness,
Large data requirements,
Lack of quantitative measures of their reliability,
Lack of clear statement of their limitations, and
Lack of clear guidance as to the conditions for their applicability.
Also, some of the models cannot be embedded with social, political, and
environmental systems.
Deficiencies of Watershed Models
Although watershed models have become increasingly more sophisticated, there is a long way
to go before they become “household” tools.
Module 9
Hydrologic Models
Model Type Example of ModelLumped parameter Snyder or Clark UHDistributed Kinematic waveEvent HEC-1, HEC-HMS, SWMM, SCS TR-20
Continuous Stanford Model, SWMM, HSPF, STORMPhysically based HEC-1, HEC-HMS, SWMM, HSPFStochastic Synthetic streamflowsNumerical Kinematic or dynamic wave modelsAnalytical Rational Method, Nash IUH
Module 9
Hydrologic Models Contd…
Models Application AreasHEC-HMS Design of drainage systems, quantifying the
effect of land use change on floodingNational Weather Service (NWS) Flood forecasting.Modular Modeling System (MMS) Water resources planning and management
worksUniversity of British Columbia (UBC) & WATFLOOD
Hydrologic simulation
Runoff-Routing model (RORB)& WBN
Flood forecasting, drainage design, and evaluating the effect of land use change
TOPMODEL & SHE Hydrologic analysisHBV Flow forecasting
Module 9
A List of Popular Hydrologic Models
Module 9
Model name/acronym Author(s)(year) RemarksStanford watershed Model (SWM)/Hydrologic Simulation Package-Fortran IV (HSPF)
Crawford and Linsley (1966), Bicknell et al. (1993)
Continuous, dynamic event or steady-statesimulator of hydrologic and hydraulic andwater quality processes
Catchment Model (CM) Dawdy and O’Donnell (1965)
Lumped, event-based runoff model
Tennessee Valley Authority (TVA) Model
Tenn. Valley Authority (1972)
Lumped, event-based runoff model
U.S. Department of Agriculture Hydrograph Laboratory(USDAHL) Model
Holtan and Lopez (1971), Holtan et al. (1974)
Event-based, process-oriented, lumped hydrograph model
U.S. Geological Survey (USGS) Model
Dawdy et al. (1970, 1978)
Process-oriented, continuous/event-basedrunoff model
Module 9
Popular Hydrologic Models
Model name/acronym Author(s)(year) RemarksUtah State University (USU) Model
Andrews et al. (1978) Process-oriented, event /continuous streamflowmodel
Purdue Model Huggins and Monke(1970)
Process-oriented, physically based, event runoff model
Antecedent Precipitation Index (API)Model
Sittner et al. (1969) Lumped, river flow forecast model
Hydrologic Engineering Center—Hydrologic ModelingSystem (HEC-HMS)
Feldman (1981), HEC (1981, 2000)
Physically-based, semi-distributed, event-based, runoff model
Streamflow Synthesis and Reservoir regulation (SSARR)Model
Rockwood (1982)U.S. Army Corps of Engineers (1987),Speers (1995)
Lumped, continuous streamflow simulation model
Module 9
Popular Hydrologic Models Contd…
Model name/acronym Author(s)(year) Remarks
National Weather service-River Forecast System (NWS-RFS)
Burnash et al. (1973a,b),Burnash (1975)
Lumped, continuous river forecast system
University of British Columbia (UBC) Model
Quick and Pipes (1977), Quick (1995)
Process-oriented, lumped parameter, continuous simulation model
Tank Model Sugawara et al. (1974) , Sugawara (1995)
Process-oriented, semi-distributed or lumped continuous simulation model
Runoff Routing Model (RORB) Laurenson (1964),Laurenson and Mein (1993, 1995)
Lumped, event-based runoff simulation model
Agricultural Runoff Model (ARM) Donigian et al. (1977) Process-oriented, lumped runoff simulation model
Module 9
Popular Hydrologic Models Contd…
Model name/acronym Author(s)(year) Remarks
Storm Water Management Model (SWMM)
Metcalf and Eddy et al. (1971),Huber and Dickinson (1988),Huber (1995)
Continuous, dynamic event or steady-statesimulator of hydrologic and hydraulic andwater quality processes
Areal Non-point Source Watershed Environment ResponseSimulation (ANSWERS)
Beasley et al. (1977),Bouraoui et al. (2002)
Event-based or continuous, lumped parameter runoff and sediment yieldsimulation model
National Hydrology Research Institute (NHRI)Model
Vandenberg (1989) Physically based, lumped parameter, continuous hydrologic simulation model
Technical Report-20 (TR-20) Model
Soil Conservation Service (1965)
Event-based, process-oriented, lumped hydrograph model
U.S. Geological Survey (USGS) Model
Dawdy et al. (1970, 1978)
Lumped parameter, event based runoff simulation model
Module 9
Model name/acronym Author(s)(year) Remarks
Physically Based Runoff Production Model (TOPMODEL)
Beven and Kirkby(1976, 1979), Beven(1995)
Physically based, distributed, continuoushydrologic simulation model
Generalized River Modeling Package—SystemeHydroloque Europeen (MIKE-SHE)
Refsgaard and Storm (1995)
Physically based, distributed, continuoushydrologic and hydraulic simulation model
ARNO(Arno River )Model Todini (1988a,b, 1996) Semidistributed, continuous rainfall-runoffsimulation model
Waterloo Flood System (WATFLOOD)
Kouwen et al. (1993),Kouwen (2000)
Process-oriented, semidistributed continuous flow simulation model
Topgraphic Kinematic Approximation and Integration (TOPIKAPI)Model
Todini (1995) Distributed, physically based, continuousrainfall-runoff simulation model
Module 9
Model name/acronym Author(s)(year) Remarks
Soil-Vegetation-Atmosphere Transfer (SVAT) Model
Ma et al. (1999),Ma and Cheng (1998)
Macroscale, lumped parameter, streamflowsimulation system
Systeme Hydrologique EuropeenTransport(SHETRAN)
Ewen et al. (2000) Physically based, distributed, water quantityand quality simulation model
Daily Conceptual Rainfall-Runoff Model (HYDROLOG)-Monash Model
Potter and McMahon (1976), Chiew and McMahon (1994)
Lumped, conceptual rainfall-runoff model
Soil Water Assessment Tool (SWAT)
Arnold et al. (1998) Distributed, conceptual, continuous simulation model
Distributed Hydrological Model (HYDROTEL)
Fortin et al. (2001a,b) Physically based, distributed, continuous hydrologic simulation model
Module 9
Lecture 2: Steps in watershed modelling
Module 9
1. Model Selection
Select model based on study objectives and watershed
characteristics, availability of data and project budget
The selection of a model is very difficult and important decision, since the
success of the analysis depends on the accuracy of the results
The selection of a model also depends upon time scale, hydrologic quantity aiming at and the processing speed of the computer at hand
Steps in Watershed Modelling
(Bedient et al., 2008)
Module 9
2. Input Data
Obtain all necessary input data-
rainfall, infiltration, physiography, landuse, channel
characteristics, streamflow, design floods and reservoir data.
1. Agricultural Data: Vegetative cover,
Land use,
Treatment, and
Fertilizer application
Module 9
Steps in Watershed Modelling Contd…
Source: CSRE, IIT Bombay
Module 9
Steps in Watershed Modelling Contd…
2. Input Data
2. Hydrometeorologic Data:
Rainfall,
Snowfall,
Temperature,
Radiation,
Humidity,
Vapor pressure,
Sunshine hours,
Wind velocity, and
Pan evaporation
Module 9
2. Input Data
Steps in Watershed Modelling Contd…
3. Pedologic Data:
Soil type, texture, and structure
Soil condition
Soil particle size, diameter, porosity
Soil moisture content and capillary pressure
Steady-state infiltration,
Saturated hydraulic conductivity, and
Antecedent moisture content
Module 9
2. Input Data
Steps in Watershed Modelling Contd…
4. Geologic Data:
Data on stratigraphy, lithology, and structural controls, Depth, and areal extent of aquifers.
For confined aquifers, hydraulic conductivity, transmissivity, storativity, compressibility, and porosity are needed.
For unconfined aquifers, data on specific yield, specific storage, hydraulic conductivity, porosity, water table, and recharge are needed.
Module 9
2. Input Data
Steps in Watershed Modelling Contd…
5. Geomorphologic Data:
Topographic maps showing:
elevation contours,
river networks,
drainage areas,
slopes and slope lengths, and
watershed area
Source: CSRE, IIT Bombay
Module 9
2. Input Data
Steps in Watershed Modelling Contd…
Source: CSRE, IIT Bombay
Module 9
2. Input Data
Steps in Watershed Modelling Contd…
Source: CSRE, IIT Bombay Module 9
2. Input Data
Steps in Watershed Modelling Contd…
Digital Elevation Model (DEM)
• Digital file that stores the elevation of the land surface in a specified grid cell size (e.g., 30 meters)
Module 9
http://www.cabnr.unr.edu/saito/intmod/docs/tootle-hydrologic-modeling.ppt.
2. Input Data
Steps in Watershed Modelling Contd…
Digital Elevation Model (DEM)
Module 9
0.002.004.006.008.0010.0012.0014.0016.0018.0020.0022.0024.0026.0028.0030.0032.0034.0036.0038.00
(Source: http://www.crwr.utexas.edu/gis/gishyd98/dhi/mike11/M11_main.htm)
Steps in Watershed Modelling Contd…
6. Hydraulic Data:
Roughness,
Flow stage,
River cross sections, and
River morphology
Module 9
2. Input Data
Steps in Watershed Modelling Contd…
7. Hydrologic Data:
Flow depth,
Streamflow discharge,
Base flow,
Interflow,
Stream-aquifer interaction,
Potential,
Water table, and
Drawdown
Design Point
1
5
6
3
2
Streamflow
http://www.cabnr.unr.edu/saito/intmod/docs/tootle-hydrologic-modeling.ppt.
Module 9
2. Input Data
Steps in Watershed Modelling Contd…
3. Evaluation & Refinement of Objectives
Evaluate and refine study objectives in terms of simulations to be performed under various watershed conditions.
4. Selection of Methodology
Choose methods for determining sub-basin hydrographs and channel routing
Module 9
Steps in Watershed Modelling Contd…
5. Calibration & Verification of Model
Model calibration involves selecting a measured set of input data
(rainfall, channel routing, landuse and so on) and measured output hydrographs
for model application.
Calibrate model using historical rainfall, streamflow and existing watershed
conditions.
Verify model using other events under different conditions while maintaining
same calibration parameters.
(Bedient et al., 2008)
Module 9
Steps in Watershed Modelling Contd…
Watershed Model Calibration
Module 9
Watershed Model(Representation)
Calibration
Model Results
Parameters
Adjust
Calibrated Model Results
Accept
OutputObservations
Steps in Watershed Modelling Contd…
6. Simulations using Model
Perform model simulations using historical or design rainfall, various conditions
of landuse and various control schemes for reservoirs, channels or diversions.
7. Sensitivity Analysis of Model
Perform sensitivity analysis on input rainfall, routing parameters and hydraulic
parameters as required.1) Precipitation 2) Soil parameters
a) Hydraulic conductivityb) Soil water holding capacity
3) Evaporation (for continuous simulation)4) Flow routing parameters (for event-based)
Module 9
Steps in Watershed Modelling Contd…
8. Model Validation
Evaluate usefulness of the model and comment on needed changes or
modifications.
Uncertainties
Precipitation
Extrapolation of point to other areas
Temporal resolution of data
Soils information
Surveys are based on site visits and then extrapolated
Routing parameters
Usually assigned based on empirical studies
Module 9
Steps in Watershed Modelling Contd…
Lecture 3: Major hydrologic models-HSPF, HEC and MIKE
Module 9
Major Hydrologic Models
HSPF (SWM)
HEC
MIKE
Module 9
HSPF is a deterministic, lumped parameter, physically based, continuous
model for simulating the water quality and quantity processes that occur in
watersheds and in a river network.
Commercial successor of the Stanford Watershed Model (SWM-IV) (Johanson et al., 1984):
Water-quality considerations
Kinematic Wave routing
Variable Time Steps
Module 9
Hydrological Simulation Program-Fortran (HSPF)
Data Requirements of HSPF:
Rainfall
Infiltration
Baseflow
Streamflow
Soils
Landuse
HSPF incorporates watershed-scale ARM (Agricultural Run-off
Management) and NPS (Non-Point Source) models into a basin-scale analysis
framework
fate and transport of pollutants in 1-D stream channels.
Module 9
HSPF Contd…
HSPF is one of the most complex hydrologic models which simulates:
Infiltration: Philip's equation, a physically based method which uses
an hourly time step
Streamflow: Chezy – Manning’s equation
HSPF can simulate temporal scales ranging from minutes to days
Due to its flexible modular design, HSPF can model systems of varying size
and complexity;
Module 9
HSPF Contd…
Stanford Watershed Model (AquaTerra, 2005)
To stream
Actual ET
Output
1
Module 9
CEPSC : interception storage capacity
LSUR : length of the overland flow plane
SLSUR : slope of the overland flow plane
NSUR : Manning's roughness of the land surface
INTFW : interflow inflow
INFILT : index to the infiltration capacity of the soil
UZSN : nominal capacity of the upper-zone storage
IRC : interflow recession constant
LZSN : nominal capacity of the lower-zone storage
LZETP : lower-zone evapotranspiration
AGWRC : basic ground-water recession rate
AGWETP : fraction of remaining potential evapotranspiration that can be satisfied from active ground-water
storage
KVARY : indication of the behavior of ground-water recession flow
DEEPFR : fraction of ground-water inflow that flows to inactive ground water
BASETP : fraction of the remaining potential evapotranspiration that can be satisfied from base flow
(Kate Flynn, U.S. Geological Survey, written commun., 2004)
Module 9
HEC Models
Module 9
HEC Models
Modeling of the rainfall-runoff process in a watershed based on watershed physiographic data
a variety of modeling options in order to compute UH for basin areas.
a variety of options for flood routing along streams.
capable of estimating parameters for calibration of each basin based on
comparison of computed data to observed data
Module 9
1. HEC-GridUtil 2.02. HEC-GeoRAS 10 (EAP) 3. HEC-GeoHMS 10 (EAP) 4. HEC-GeoEFM 1.0 5. HEC-SSP 2.0 6. SnoTel 1.2 Plugin7. HEC-HMS 3.5 8. HEC-FDA 1.2.5a
9. HEC-DSSVue 2.0.1 10. HEC-RAS 4.1 11. HEC-DSS Excel Add-In 12. HEC-GeoDozer 1.0 13. HEC-EFM 2.0 14. HEC-EFM Plotter 1.0 15. HEC-ResSim 3.0 16. HEC-RPT 1.1
HEC-GridUtil is designed to provide viewing, processing, and analysis capabilities for gridded data sets stored in HEC-DSS format (Hydrologic Engineering Center's Data Storage System).
http://www.hec.usace.army.mil/software/hec-gridutil/documentation.html
HEC-GridUtil 2.0
Module 9
HEC-GeoRAS
Module 9
GIS extension a set of procedures, tools, and utilities for the preparation of GIS
data for import into HEC-RAS and generation of GIS data from RAS output.
HEC-GeoRAS 10 (EAP)
• ArcGIS w/ extensions 3D & Spatial Analyst HEC-GeoHMS HEC-GeoRAS
• HEC-RAS– Simulates water surface profile of a stream reach
Module 9
Data Requirements
• Triangular Irregular Network (TIN)
• DEM (high resolution)– use stds2dem.exe if
downloading from USGS
• Land Use / Land Cover– Manning’s Coefficient
Module 9
CRWR image, Texas University
(Source: “GIS – Employing HEC-GeoRAS”, Brad Endres, 2003)
Major Functions of GeoRAS
• Interface between ArcView and HEC-RAS• Functions:
– PreRAS Menu - prepares Geometry Data necessary for HEC-RAS modeling– GeoRAS_Util Menu – creates a table of Manning’s n value from land use
shapefile– PostRAS Menu – reads RAS import file; delineates flood plain; creates
Velocity and Depth TINs
Module 9
Demonstration of Capabilities
• Load TIN
• Create Contour Lines
Module 9
3-D Scene
3-D Scene
Demonstration of Capabilities Contd…
• Create Stream Centerline
• Create Banks Theme
• Create Flow Path Centerlines
• Create Cross Section Cut Lines
• Add/Create Land Use Theme
• Generate RAS Import File
Module 9
Stream Centerline
Right BankFlow Path Centerlines
Land Use Themes
Cross SectionCut Lines
Module 9
Demonstration of Capabilities Contd…
Generate RAS GIS import file Open HEC-RAS and import RAS GIS file Complete Geometry, Hydraulic, & Flow Data Run Analysis Generate RAS Export file
Module 9
Demonstration of Capabilities Contd…
RAS GIS import file
Module 9
Demonstration of Capabilities Contd…
RAS GIS export file
Module 9
Demonstration of Capabilities Contd…
• New GIS data• PostRAS features
Water Surface TIN
Floodplain Delineation – polygon & grid
Velocity TIN
Velocity Grid
Module 9
Demonstration of Capabilities Contd…
Floodplain Delineation (3-D Scene)
Module 9
Demonstration of Capabilities Contd…
Depth Grid (Darker = Deeper) Velocity Grid (Darker = Faster)
Module 9
Employing ArcView, GeoRAS, and RAS for Main Channel Depth Analysis (1968)
Module 9
PreRAS PostRAS
13.5 ft
Employing ArcView, GeoRAS, and RAS for Main Channel Depth Analysis (1988)
PreRAS PostRAS
21.0 ft
Module 9
Overall Benefits
Elevation data is more accurate with TIN files
Better representation of channel bottom
Rapid preparation of geometry data (point and click)
Precision of GIS data increases precision of geometry data
Efficient data transport via import/export files
Velocity grid
Depth grid
Module 9
Floodplain maps can be made faster
• several flow scenarios
Both steady & unsteady flow analysis
GIS tools aid engineering analysis• Automated calculation of functions (Energy Equation)• Structural validation of hydraulic control features• Voluminous data on World Wide Web
Makes data into visual event – easier for human brain to process!
Module 9
Overall Benefits Contd…
Overall Drawbacks
Time required to learn several software packages
Non-availability of TIN or high resolution data
Estimation of Manning’s Coefficient• Few LU/LC files have this as attribute data
Velocity distribution data may not be calculated• HEC-RAS export file without velocity data means no velocity TIN or
grid
Module 9
HEC-HMS
HEC-HMS simulates rainfall-runoff for the watershed
(Source: ftp://ftp.crwr.utexas.edu) Module 9
HEC-HMS Background
Purpose of HEC-HMS
Improved User Interface, Graphics, and Reporting
Improved Hydrologic Computations
Integration of Related Hydrologic Capabilities
Importance of HEC-HMS
Foundation for Future Hydrologic Software
Replacement for HEC-1
Module 9
Ease of Use projects divided into three components
user can run projects with different parameters instead of creating new
projects
hydrologic data stored as DSS files
capable of handling NEXRAD-rainfall data and gridded precipitation
Converts HEC-1 files into HMS files
Module 9
Improvements over HEC-1
HEC-1 EXERCISE PROBLEM
A small undeveloped watershed has the parameters listed in the following tables. A unit hydrograph and Muskingum routing coefficients are known for subbasin 3, shown in Fig.1(a). TC and R values for subbasins 1 and 2 and associated SCS curve numbers (CN) are provided as shown. A 5-hr rainfall hyetograph in in./hr is shown in Fig.1(b) for a storm event that occurred on July 26, 2011. Assume that the rain fell uniformly over the watershed. Use the information given to develop a HEC-1 input data set to model this storm. Run the model to determine the predicted outflow at point B. SUBBASIN NUMBER
TC (hr)
R (hr )
SCS CURVE NUMBER
% IMPERVIOUS (% )
AREA (mi2)
1 2.5 5.5 66 0 2.5 2 2.8 7.5 58 0 2.7 3 -- -- 58 0 3.3
UH FOR SUBBASIN 3:
TIME (hr) 0 1 2 3 4 5 6 7
U (cfs) 0 200 400 600 450 300 150 0
(Bedient et al., 2008)
Module 9
Fig.1(a) Fig.1(b)Muskingum coefficients: x = 0.15, K = 3 hr, Area = 3.3 sq mi
Module 9
Example Problem Contd…
ID ****ID ****ID ****ID ****IT 60 60 25-Jul-07 1200 100IO 4KK SUB1KMPI 0.2 1.5 2 1 0.5BA 2.5LS 66 0UC 2.5 5.5KK SUB2KMBA 2.7LS 58 0UC 2.8 7.5KK AKMHC 2
KMRM 1 3 0.15KK SUB3KMBA 3.3LS 58 0UI 0 200 400 600 450 300 150
MUSKINGUM ROUTING FROM A TO B
RUNOFF FROM SUBBASIN 3
KKA TO B
EXAMPLE PROBLEM
HEC-1 INPUT DATA SET
RUNOFF FROM SUBBASIN 1
RUNOFF FROM SUBBASIN 2
COMBINE RUNOFF FROM SUB 1 WITH RUNOFF FROM SUB 2 AT A
Solution : The input data set is as follows:
Module 9
Using HEC-HMS Contd…
Three components
Basin model - contains the elements of the basin, their connectivity, and
runoff parameters ( It will be discussed in detail later)
Meteorologic Model - contains the rainfall and evapotranspiration data
Control Specifications - contains the start/stop timing and calculation
intervals for the run
Module 9
Project Definition
It may contain several basin models, meteorological models, and control specifications
It is possible to select a variety of combinations of the three models in order to see the effects of changing parameters on one sub-basin
Module 9
Basin Model
GUI supported
Click on elements from left and drag into basin area
Works well with GIS imported files
Actual locations of elements do not matter, just connectivity and runoff parameters
Module 9
1. Basin Model Elements
• subbasins- contains data for subbasins (losses, UH transform, and baseflow)
• reaches- connects elements together and contains flood routing data
• junctions- connection point between elements
• reservoirs- stores runoff and releases runoff at a specified rate (storage-discharge relation)
Module 9
1. Basin Model Elements Contd…
• sinks- has an inflow but no outflow
• sources- has an outflow but no inflow
• diversions- diverts a specified amount of runoff to an element based on a rating curve - used for detention storage elements or overflows
Module 9
a) Loss rate
b) Transform
c) Baseflow methods
Module 9
2. Basin Model Parameters
2a) Abstractions (Losses)
1. Interception Storage
2. Depression Storage
3. Surface Storage
4. Evaporation
5. Infiltration
6. Interflow
7. Groundwater and Base Flow
Module 9
2. Basin Model Parameters Contd…
1. Unit Hydrograph2. Distributed Runoff3. Grid-Based Transformation
Methods:a. Clark b. Snyder c. SCSd. Input Ordinates e. ModClarkf. Kinematic Wave
2b) Transformation
2c) Baseflow Options
a. recession
b. constant monthly
c. linear reservoir
d. no base flow
Module 9
2. Basin Model Parameters Contd…
Stream Flow Routing
Simulates Movement of Flood Wave Through Stream Reach
Accounts for Storage and Flow Resistance
Allows modeling of a watershed with sub-basins
Module 9
a) Simple Lag
b) Modified Puls
c) Muskingum
d) Muskingum Cunge
e) Kinematic Wave
Reach Routing
Hydraulic Methods - Uses partial form of St Venant Equations
Kinematic Wave Method
Muskingum-Cunge Method
Hydrologic Methods
Muskingum Method
Storage Method (Modified Puls)
Lag Method
Module 9
Methods for Stream Flow Routing
Developed Outside HEC-HMS
Storage Specification Alternatives:
Storage versus Discharge
Storage versus Elevation
Surface Area versus Elevation
Discharge Specification Alternatives:
Spillways, Low-Level Outlets, Pumps
Dam Safety: Embankment Overflow, Dam Breach
Module 9
Reservoir Routing
Reservoirs
Q (cfs)
I=Q
time
Q (cfs)
Inflow
Outflow
I - Q = dS dt
Level Pool Reservoir Q (weir flow)
Q (orifice flow)
I
SH
S = f(Q) Q = f(H)
Orifice flow:
Q = C * 2gH
Q
I
I
Weir Flow: Q = CLH3/2
Q
Pond storage with
outflow pipe
Orifice flow
Weir flows
Inflow and Outflow
Module 9
Initial Conditions to be considered
Inflow = Outflow
Initial Storage Values
Initial Outflow
Initial Elevation
Elevation Data relates to both Storage/Area and Discharge
HEC-1 Routing routines with initial conditions and elevation data
can be imported as Reservoir Elements
Module 9
Reservoir Data Input
Module 9
Reservoir Data Input Window
User selects:
1. Basin model
2. Meteorologic model
3. Control ID for the
HMS run
Running a project
Module 9
To view the results:
• right-click on any basin element, results will be for that point
Display of results:
• hydrograph- graphs outflow vs. time
• summary table- gives the peak flow and time of peak
• time-series table- tabular form of outflow vs. time
Comparing computed and actual results:
• plot observed data on the same hydrograph to by selecting a discharge
gage for an element
Module 9
Viewing Results
hydrograph
Module 9
Viewing Results Contd…
HEC-HMS Output
1. Tables
Summary
Detailed (Time Series)
2. Hyetograph Plots
3. Sub-Basin Hydrograph Plots
4. Routed Hydrograph Plots
5. Combined Hydrograph Plots
6. Recorded Hydrographs - comparison
Module 9
Summary table
Time series table Module 9
Viewing Results
Sub-Basin Plots
Runoff Hydrograph
Hyetograph
Abstractions
Base Flow
Module 9
Viewing Results Contd…
Junction Plots
Module 9
a. Tributary Hydrographs
b. Combined Hydrograph
c. Recorded Hydrograph
Viewing Results Contd…
Lecture 4: MIKE models
Module 9
a) Flood Management
• MIKE 11 For Analyzing Open Channel Flow
• MIKE 21 For Analyzing surface complicated overflow
b) River Basin Management
• MIKE BASIN
c) Hydrological Cycle
• MIKE SHE
d) Urban Drainage
• MOUSE For Analyzing Urban Sewage
MIKE Models
Module 9
1 Dimensional
• MIKE 11
2 Dimensional
• MIKE 21
3 Dimensional
• MIKE SHE
• MIKE 3
Module 9
MIKE Hydrological and Hydrodynamic Models
MIKE Zero-fication!
MIKE ZERO
MIKE 3
MIKE 11
MIKE SHE
MIKE 21
Module 9
Module 9
ArcMap with MIKE ZERO
Module 9
MIKE 11 GIS
Fully integrated GIS based flood modelling Developed in ArcView GIS Pre-processing:
Post-processing:
Analysis with other GIS data
Floodplain schematization
•Flood depth maps •Comparison maps•Duration maps
Module 9
Mike 11 Modules HD : hydrodynamic - simulation of unsteady flow in a network of
open channels. Result is time series of discharges and water levels;
AD : advection dispersion WQ : water quality
Saint Venant equations (1D)
continuity equation (mass conservation)
momentum equation (fluid momentum conservation)
Assumptions
water is incompresible and homogeneous
bottom slope is small
flow everywhere is paralel to the bottom ( i.e. wave lengths are large
compared with water depths)
Module 9
Open channel flow
Discretization - branches
Module 9
IHE, 2002
Module 9IHE, 2002
Discretization – branches Contd...
Discretization - cross sections
Required at representative locations throughout the branches of the river
Must accurately represent the flow changes, bed slope, shape, flow
resistance characteristics
Module 9IHE, 2002
Friction formulas
Chezy
Manning
For each section a curve is made with wetted area, conveyance factor,
hydraulic radius as a function of water level
Module 9
h
R
Discretization – cross sections Contd...
IHE, 2002
Module 9
Typical Model Schematisation
Minor river
Major River
Floodplains
Spill channel
Spill channel
(Source: http://www.crwr.utexas.edu/gis/gishyd98/dhi/mike11/M11_main.htm)
Mike 11 main menu
Module 9
Extraction from DEM
Module 9(Source: http://www.crwr.utexas.edu/gis/gishyd98/dhi/mike11/M11_main.htm)
Import to MIKE 11
Module 9(Source: http://www.crwr.utexas.edu/gis/gishyd98/dhi/mike11/M11_main.htm)
Menus and input files editors
Module 9
Network editor
Module 9
Tabular view
Graphical view
Module 9
River network -branches connection
Network editor Contd...
River network -structures
Module 9
Network editor Contd...
Parameter file editor
Module 9
Parameter file - coefficients
Module 9
h1
Time step n+1/2
Time step n
Time Time step n+1
i i+1 i-1
Space
h3
h5
h7
4
6
Q
Centerpoint
inii
nii
nii QhQq
tA
xQ δγβα =++⇒=
∂∂
+∂∂ +
+++
−1
111
1
inii
nii
nii hQh
ARCQgQ
xhgA
xA
Q
tQ δγβα
α=++=+
∂∂
+∂
∂
+∂∂ +
+++
−1
111
12
2
0
2
Because of its numerical limitations, MIKE 11 cannot model the supercritical
flow downstream of the weir.
For the low-flow case, the downstream water level is over-estimated by
a factor of 8 .This high tailwater, impacts on the flow conditions on the
weir, causing a significant error in the upstream water level.
The incorrect tailwater has less impact for the high-flow case. There is
still significant error in the predictions across the weir, but the upstream
water level is almost correct.
Limitations of MIKE 11
Module 9
Advanced integrated hydrological modeling system
Simulates water flow in the entire land based phase of the hydrological cycle
rainfall to river flow, via various flow processes such as,
overland flow,
infiltration into soils,
evapotranspiration from vegetation, and
groundwater flow
MIKE SHE
Module 9
Integrated:
Fully dynamic exchange of water between all major hydrological components is
included, e.g. surface water, soil water and groundwater
Physically based:
It solves basic equations governing the major flow processes within the study area
Fully distributed:
The spatial and temporal variation of meteorological, hydrological, geological and
hydrogeological data
Modular:
The modular architecture allows user only to focus on the processes, which are
important for the study
MIKE SHE Features
Module 9
Hydrological Processes simulated by MIKE SHEModule 9
Schematic view of the process in MIKE SHE, including the available numeric engines for each process.
The arrows show the available exchange pathways for water between the process models.
(V.P. Singh & D.K. Frevert, 199
Module 9
Flow System Inputs and Outputs
Module 9
Reservoir nameReservoir IDInitial water levelPriority of inflow connectionsPriority of down-stream user(s)Down-streams user(s) reduction factorDown-stream user(s) loss factorRule curvesHeight, Volume, AreaPrecipitationEvaporation
Hydrodynamic Integrated model with 1D and 2D
The result of either 1D or 2D model can be transferred as input
of another model
This model simulates simultaneously the flow in the sewer, the
drainage system and the surface flow
Module 9
MIKE FLOOD
Conceptual Representations
Rainfall Runoff
Surface RunoffOverflow
Re-inflow after flood
Flow-capacity excess
Sewage/Rainfall waterSewage/Rainfall water
Module 9
Lecture 5: Urban Flood Risk Mapping using MOUSE, MIKE 21 and MIKE FLOOD
Module 9
MIKE FLOOD
Set up
Coupling
Preprocessing for MOUSE
Run Model
Check the result and evaluation
Set up
Validation
MIKE21 Model
MOUSE Model
Set up
ValidationPreparationGIS
Rainfall Analysis
Module 9
Urban Drainage Network
Import of drainage network into MOUSE
Setting up Urban Drainage model with MOUSE
Validation
Module 9
Import Ascii data-set as bathymetry
Setting up Urban Bathymetry with MIKE21
Validation
Option Value
Time Step1secMax Cr=0.4
Grid Size 10m
Urban Drainage Modelling
Module 9
Setting up MIKE Flood model
Coupling : Link MOUSE Manholes to MIKE21
Preprocessing for MOUSE Model
Running Model
Check the results
(Source: www.hydroasia.org)
MIKE FLOOD Modeling
Module 9
Catchment boundary
Gajwa-WWTP
Cityhall
Ganseok st.Juan st.
Incheon gyo
Pump Station
Flooded area
Results Analysis
(Source: www.hydroasia.org) Module 9
Comparing the Result (Modeled VS Reported)
Hwasu Reservoir basin
Around Dohwa and Juan metro station
Seoknam Canal Basin
Gajwa girls juniorhigh
Sipjeong & Ganseok dong
Seoknam Canal
Hwasu Reservoir basin
Around Dohwa and Juan metro station
Seoknam Canal Basin
Gajwa girls juniorhigh
Sipjeong & Ganseok dong
Seoknam Canal
(Source: www.hydroasia.org) Module 9
Highlights in the Module
Basic elements of hydrologic simulation model Equations govern the hydrologic processes,Maps define the study area andDatabase numerically describe the study area and model parameters.
Strengths of Watershed Models Diversity of the current generation of models
Comprehensive Nature
Reasonable modeling of physical phenomena
Distributed in Space and Time
Multi-disciplinary nature
Module 9
Deficiencies of Watershed Models Lack of user-friendliness,
Large data requirements,
Lack of quantitative measures of their reliability,
Lack of clear statement of their limitations, and
Lack of clear guidance as to the conditions for their applicability.
Steps in Watershed Modelling Model Selection
Input Data
Agricultural Data, Hydrometeorologic Data, Pedologic Data, Geologic
Data, Geomorphologic Data, Hydraulic Data, Hydrologic Data
Highlights in the Module Contd…
Module 9
Evaluation & Refinement of Objectives
Selection of Methodology
Calibration & Verification of Model
Simulations using Model
Sensitivity Analysis of Model
Model Validation
Major Hydrologic Models
HSPF (SWM)
HEC
MIKE
Module 9
Highlights in the Module Contd…
Text/References
1.Abraham, K.R.,Dash, S.K., and Mohanty, U.C.1996. Simulation of monsoon
circulation and cyclones with different types of orography; Mausam, 47, 235-
248.
2.Bedient et al. “Hydrology and Floodplain Analysis”, 2008.
3.Bhalme, H.N. and Jadhav, S.K. 1984. The southern oscillation and its
relation to the monsoon rainfall. J.Climatol., 4, 509-520.
4.Bras, R. L., and Rodriguez-Iturbe. 1994. Random Functions and Hydrology,
Dover Publications, New York.
5.Chow, V. T., D. R. Maidment, and L. W. Mays. “Applied Hydrology”, McGraw
Hill International Editions.
6.Haan, C. T.. 2002. “Statistical Methods in Hydrology”, 2nd ed., Blackwell
Publishing, Ames, IA.
Text/References Contd…
7. Hoskings, J. R. M. and J. R. Wallis. 1997. “Regional Frequency
Analysis, An Approach Based on L-Moments”, Cambridge University
Press, New York.
8. Subramanya K, “Engineering Hydrology”, Tata McGraw-Hill.
9. Viessman Jr., W. and G. L. Lewis. “Introduction to Hydrology”, 4th
ed., Harper-Collins, New York, 1996.
10.Vijay P. Singh, F. and David A. Woolhiser, M. 2002. Mathematical Modeling
of Watershed Hydrology, Journal of Hydrologic Engineering, Vol. 7, No.
4, July 1, 2002.
11.www.climateofindia.pbworks.com/
12.http://www.crwr.utexas.edu/gis/gishyd98/dhi/mikeshe/Mshemain.htm
Related study materials available through NPTEL
http://nptel.iitm.ac.in/courses/105104029/
1. Advanced Hydrology (Video Course) by Prof. Ashu Jain, IIT Kanpur
2. Stochastic Hydrology (Video Course) by Prof. P.P. Mujumdar, IISc Bangalore
http://nptel.iitm.ac.in/courses/105108079/
3. Probability Methods in Civil Engineering (Video Course) by Dr. Rajib Maity, IIT Kharagpur
http://nptel.iitm.ac.in/courses/105105045/