Assignment 2

Assignment 2 | group 2 | DR Balqis Rehan

Types of extreme value models
Group 2
 Ahmed fuad zainudin GS: 51033
 Eddy Herman Binu Sharu GS:51229
 Mohamed Haibe Madar GS: 51593
 Mohamed Muhumed Yousuf GS: 53450
 Types of extreme value models:
extreme events:
 Extreme value analysis aims to quantify the behaviour of a process at extremely high or low levels. From there one can try to
estimate the probabilities of events occurring at levels more extreme than already witnessed.
 Extreme events
– Floods
– Droughts
 Magnitude of extreme events is related to their frequency of occurrence

 The objective of frequency analysis is to relate the magnitude of events to their frequency of occurrence through probability
distribution
 It is assumed the events (data) are independent and come from identical distribution
 Extreme values….
 Extreme value series
– Includes largest or smallest values in equal intervals
 Annual series: interval = 1 year
 Annual maximum series: largest values
 Annual minimum series : smallest values
 Probability distributions
 The probability distribution describes the range of possible values that a random variable can
attain and the probability that the value of the random variable is within any (measurable)
subset of that range.
A probability distribution gives important information about the data:-
 It tells us how the values are changing, whether they are bunched together or spread out.
 If they are symmetrically disposed on the X-axis or not.

 Distribution also tells the relative frequency or proportion of various X values in the population
in the same way that a histogram gives information about a sample.
 Probability distributions ……..
 Commonly used distributions in hydrology are:

 Normal family:
– Normal, lognormal, lognormal-III

 Generalized extreme value family

– EV1 (Gumbel), GEV, and EVIII (Weibull)

 Exponential/Pearson type family

– Exponential, Pearson type III, Log-Pearson type III.
Normal Distribution

 When a hydrologic variable, integrated over a large time period is used in analysis
 the variable is usually expected to follow a normal distribution.
 The normal distribution has a symmetrical bell-shaped probability density function, there
are two main parameters in this type which are mean (µ) and standard deviation (σ).
 Hydrological events follow this
Form (are precipitation,
annual average streamflow.)
lognormal distribution
 The logarithm of such random variable can be expected to follow a normal distribution Y =
exp(X) or X = log(y).
 This approach represents a process that is the resultant of multiplicative product of so many
small effect of each one is positive.
 If the pdf of X is skewed, it’s not normally distributed
 If the pdf of Y = log (X) is normally distributed, then X is said to be log normally distributed.

1  ( y   y )2 
f ( x)  exp   x  0, and y  log x
x 2  2 y 
2


(Hydraulic conductivity, distribution of raindrop sizes in storm follow

lognormal distribution.)
Generalized extreme value family:
 This usually talks about three parameters location(µ) , scale( σ)and shape parameters (k).
 Extreme values – maximum or minimum values of sets of data
 Annual maximum discharge, annual minimum discharge
 When the number of selected extreme values is large, the distribution converges to one of the three forms of
EV distributions called Type I, II and III.
 Contains three distribution types , GUMBLE, FRECHET AND WEIBUL

1  x u  x  u 
K=0 ( EV TYPE 1) f ( x)  exp   exp  
     
K<0 ( EV TYPE 11)
6sx
K>0 (EV TYPE 3)  u  x  0.5772
Distribution of annual maximum streamflow follows an EV1 distribution

 Extreme value (EV) distributions
 Extreme values – maximum or minimum values of sets of data
 Annual maximum discharge, annual minimum discharge
 When the number of selected extreme values is large, the distribution converges to one
of the three forms of EV distributions called Type I, II and III

11
EV type I distribution “Gumble distribution”
If M1, M2…, Mn be a set of daily rainfall or streamflow, and let X = max(Mi) be the maximum
for the year. If Mi are independent and identically distributed, then for large n, X has an extreme
value type I or Gumbel distribution.
Gumble…

 This distribution is widely used for frequency analaysis of floods.

 Maximum rainfall and etc.
EV type III distribution
 If Wi are the minimum streamflows in different days of the year, let X = min(Wi) be the
smallest. X can be described by the EV type III or Weibull distribution.

 k  x 
k 1
  x k 
f ( x)     exp     x  0;  , k  0
         

Distribution of low flows (eg. 7-day min flow) follows EV3

distribution.
Exponential distribution
 Poisson process – a stochastic process in
which the number of events occurring in
two disjoint subintervals are independent
random variables.
 In hydrology, the interarrival time (time
between stochastic hydrologic events) is
described by exponential distribution

 x 1
f ( x )  e x  0;  
x
Interarrival times of polluted runoffs, rainfall intensities, etc. are
described by exponential distribution.
Gamma Distribution/Pearson Type III distribution
 The time taken for a number of events (b) in
a Poisson process is described by the
gamma distribution
 Gamma distribution – a distribution of sum
of b independent and identical
exponentially distributed random variables.
b x b 1e  x
f ( x)  x  0;   gamma function
( b )

Skewed distributions (eg. hydraulic conductivity) can be

represented using gamma without log transformation.
 Log-Pearson Type III
 If log X follows a Person Type III distribution, then X is said to have a
log-Pearson Type III distribution.

b ( y   ) b 1 e   ( y  )
f ( x)  y  log x  
( b )
 Normal probability plot
 Steps
1. Rank the data from largest (m = 1) to smallest (m = n)
2. Assign plotting position to the data
1. Plotting position – an estimate of exccedance probability
2. Use p = (m-3/8)/(n + 0.15)
3. Find the standard normal variable z corresponding to the
plotting position (use -NORMSINV (.) in Excel)
4. Plot the data against z
 If the data falls on a straight line, the data comes from
a normal distributionI

18
 Normal Probability Plot
600

500
Data

Q (1000 cfs)
400 Normal

300
200

100

0
-3 -2 -1 0 1 2 3
Standard normal variable (z)

Annual maximum flows for Colorado River near Austin, TX

The pink line you see on the plot is xT for T = 2, 5, 10, 25, 50, 100, 500 derived using
the frequency factor technique for normal distribution.
19
 EV1 probability plot
 Steps
1. Sort the data from largest to smallest
2. Assign plotting position using Gringorten formula pi = (m – 0.44)/(n
+ 0.12)
3. Calculate reduced variate yi = -ln(-ln(1-pi))
4. Plot sorted data against yi

 If the data falls on a straight line, the data comes from an

EV1 distribution

20
 EV1 probability plot
600

500

Data
400 EV1

Q (1000 cfs)
300

200

100

0
-2 -1 0 1 2 3 4 5 6 7
EV1 reduced variate

Annual maximum flows for Colorado River near Austin, TX

The pink line you see on the plot is xT for T = 2, 5, 10, 25, 50, 100, 500 derived using
the frequency factor technique for EV1 distribution.
21