| Author | Warain, Cecilia Tanudtanud |
| Call Number | AIT Thesis no. WA-84-19 |
| Subject(s) | Hydrology--Mathematical models
|
| Note | A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering, School of Engineering and Technology |
| Publisher | Asian Institute of Technology |
| Abstract | Numerous underlying probability density functions have been used to describe sequences of peak flood flow rates and flood volumes. As already noted , the Gumbel or type I extreme value distribution has been commonly
used in the theoretical treatment of flood events, but because of the limitation in application it is appropriate to consider also the practicability of the other two extreme value asymptotic distribution (EV2 a nd EV3).
Four different parameter estimation techniques of the general extreme value distribution have been evaluated. For the method of moments, a skewness
correction factor is being proposed in this study which is applicable to the general extreme value distribution. An extensive mathematical development for the method of maximum likelihood and sextile method in which Jenkinson
has only provided some tabular results is done. Polynomial curve fitting were performed for the different relationship of the parameter and properties
on these methods. With the advent of electronic computer usage nowadays the
approximating polynomials provided in this study give convenience to the users to yield faster results. The method of Monte Carlo of simulation was done from which the results obtained would b e the basis of the determination of the best method or best methods.
There are two cases considered in the evaluation that is when the variates belong to EV2 distribution and the other one is for the EV3 distribution . The
results obtained from this study indicate that for EV2 distribution the method of maximum likelihood is better for quantile determination at lower probability levels and the method of moments with the proposed skewness correction is best for higher probability levels. In the case of EV3 distribution sextile method is recommended for low probability levels and the method of moments for high
probability levels. |
| Year | 1983 |
| Type | Thesis |
| School | School of Engineering and Technology |
| Department | Department of Civil and Infrastucture Engineering (DCIE) |
| Academic Program/FoS | Water Resources Research Engineering (WA) |
| Chairperson(s) | Hoshi, Kiyoshi
|
| Examination Committee(s) | Yoganarashimhan , G.N.
|
| Scholarship Donor(s) | Carl Duisberg Gesselschaft |
| Degree | Thesis (M.Eng.) - Asian Institute of Technology, 1983 |