AI SearchBest fitting distribution function for floods in the Upper Indus Basin (Pakistan)
Call Number: AIT Thesis no.AE-92-50 Material type:
SeriesSeries: Asian Institute of Technology. Thesis ; no. AE-92-50Publication details: Bangkok : Asian Institute of Technology, 1992Description: 71 leaves : illSubject(s): Online resources: Dissertation note: Thesis (M.Eng.) - Asian Institute of Technology Summary: The results offlood frequency analysis can be used for the design of dams, culverts, flood control structures and to determine the economic value of projects etc,. A variety ofdistributions have been used in hydrology in fitting the hydrologic random variables of annual stream flows. So, the selection of a distribution plays an important role in the design of structure and its economic condition. In Pakistan the use of the Gumbel (EYl) distribution is a common practice. However there is no investigation on the proper use of statistical distributions for annual maximum flood frequency analysis in Pakistan. Therefore the upper Indus basin which is the largest one in Pakistan was selected for this study. A flood frequency analysis of the upper Indus basin was carried out for 20 unregulated stream flow stations located on various rivers with each data set having at least 20 observations. The Wald-Wolfowitz test of independence and Grubbs & Beck test for detection of outliers were applied. Extreme value type 1 (EYl), General extreme value (GEY), Lognormal 2 (LN2), and Pearson Type 3 (P3) distributions were used as the parent models. The Probability Weighted Moments (PWM), and Maximum Likelihood (ML) methods were taken for parameter estimation. Commonly considered criteria to evaluate the performance of distributions such as statistical measures, goodness of fit test, and probability plots using appropriate probability papers were employed. The results of the study conclude that for flood frequency analyses of the upper Indus basin, GEY (EY2 and EY3) distribution should be applied because it gives a good fit on GEY probability paper with less RMSE, BIAS, and K-S statistics. Moreover, for this distribution, the PWM method should be used because it is less biased and more efficient. The EYl, LN2, P3 distributions can not satisfactorily represent the flood data in a large number of cases. As such, they are not suitable for flood frequency analyses of the upper Indus basin.
| Cover image | Item type | Current library | Home library | Collection | Shelving location | Call number | Materials specified | Vol info | URL | Copy number | Status | Notes | Date due | Barcode | Item holds | Item hold queue priority | Course reserves | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
22-AIT Thesis (Replacement)
|
Asian Institute of Technology Library AIT Publications | AIT Thesis no.AE-92-50 (Browse shelf(Opens below)) | 2 | Available | 30050120577829 | |||||||||||||
40-Archives
|
Asian Institute of Technology Library Archives | AIT Thesis no.AE-92-50 (Browse shelf(Opens below)) | Available | 30050120362172 |
Thesis (M.Eng.) - Asian Institute of Technology
A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering.
The results offlood frequency analysis can be used for the design of dams, culverts, flood control structures and to determine the economic value of projects etc,. A variety ofdistributions have been used in hydrology in fitting the hydrologic random variables of annual stream flows. So, the selection of a distribution plays an important role in the design of structure and its economic condition. In Pakistan the use of the Gumbel (EYl) distribution is a common practice. However there is no investigation on the proper use of statistical distributions for annual maximum flood frequency analysis in Pakistan. Therefore the upper Indus basin which is the largest one in Pakistan was selected for this study. A flood frequency analysis of the upper Indus basin was carried out for 20 unregulated stream flow stations located on various rivers with each data set having at least 20 observations. The Wald-Wolfowitz test of independence and Grubbs & Beck test for detection of outliers were applied. Extreme value type 1 (EYl), General extreme value (GEY), Lognormal 2 (LN2), and Pearson Type 3 (P3) distributions were used as the parent models. The Probability Weighted Moments (PWM), and Maximum Likelihood (ML) methods were taken for parameter estimation. Commonly considered criteria to evaluate the performance of distributions such as statistical measures, goodness of fit test, and probability plots using appropriate probability papers were employed. The results of the study conclude that for flood frequency analyses of the upper Indus basin, GEY (EY2 and EY3) distribution should be applied because it gives a good fit on GEY probability paper with less RMSE, BIAS, and K-S statistics. Moreover, for this distribution, the PWM method should be used because it is less biased and more efficient. The EYl, LN2, P3 distributions can not satisfactorily represent the flood data in a large number of cases. As such, they are not suitable for flood frequency analyses of the upper Indus basin.
There are no comments on this title.