Pakistan Science Abstracts
Article details & metrics
No Detail Found!!
A comparison of normal and log-normal(2) distributions using three estimation methods on annual flood peaks data of river Indus at Kotri barrage.
Author(s):
1. Abdul Ghani Memon: Department of Statistics, Shah Abdul Latif University, Khairpur, Pakistan
Abstract:
Normal and Log-Normal LN(2) distributions are fitted, compared and tested using X2-test, Simirnov - Kolmorgrov (S-K) test, Probability Plot Correlation Co-efficient (PPCC), Co-efficient of Skewness (Cs) and Normality tests on the data of flood peaks at Kotri Barrage, from 1901-1999. Method of Moments (MOM), Maximum Likelihood Method (MLM) and Probability Weighted Moments (PWM) are used for estimating the quantiles and their Standard Errors (S.E`s) for each distribution. Confidence bands are also drawn which guide the reliability of the frequency curve. In this Paper it is shown that Normal distribution is not a proper fit (although its quantiles and S.E`s are smaller) while LN(2) appears to be appropriate. MOM is found to be the most efficient method for LN(2). The Quantile estimate for 100 years for LN(2) by MOM is 0.95 million cusecs, which is 8% higher than its designed capacity. Hence, there is need to take appropriate measures to save the barrage from any damage by the occurrence of above predicted flood flow.
Page(s): 37-46
DOI: DOI not available
Published: Journal: Pakistan Journal of Statistics, Volume: 23, Issue: 1, Year: 2007
Keywords:
Keywords are not available for this article.
References:
References are not available for this document.
Citations
Citations are not available for this document.
0

Citations

0

Downloads

10

Views