Author(s):
1. Muhammad Ali:
Department of Computer Science, Sindh Madrasatul Islam University,Aiwan-e-Tijarat Road, Karachi, Pakistan
2. Muhammad Jawed Iqbal:
Institute of Space and Planetary Astrophysics, University of Karachi, Karachi, Pakistan
3. Zohaib Aziz:
Department of Statistics, Federal Urdu University of Arts, Science and Technology,
Karachi, Pakistan
Abstract:
The consequences of statistical modeling of extreme rainfall are pivotal for civil engineering and planning division in order to instill the capability of stmctures of building that can withstand the extreme situation. Yearly maximum rainfall data of Karachi, Badin, Chhor from 1961-2010, and Rohri from 19712010 have been used in this study. The method of Maximum Likelihood (ML) and Bayesian techniques have been implemented to estimate the parameters of Generalized Extreme Value (GEV) distribution and also to compute return levels against sundry return periods. Non-informative priors are used to get the posterior densities. To gauge and compare the results of the above mentioned methods, acceptance rates and forecasting errors have been used as Goodness of Fit (GoF) test. Though both the methods are applicable, but the GoF test highlights that ML method is slightly better than Bayesian for observing the annual maximum rainfall in Sindh province of Pakistan.
Page(s):
225-235
DOI:
DOI not available
Published:
Journal: Proceedings of the Pakistan Academy of Sciences: A. Physical and Computational Sciences, Volume: 53, Issue: 2, Year: 2016
Keywords:
return period
,
return levels
,
extreme value modeling
,
goodness of fit test
,
Yearly maximum rainfall
References:
References are not available for this document.
Citations
Citations are not available for this document.