Application of Bayesian Monte Carlo Technique to Calculate Extreme Rainfall over Sindh Province in Comparison with Maximum Likelihood Method | Proceedings of the Pakistan Academy of Sciences: A. Physical and Computational Sciences 2016

Author(s)

Muhammad Ali Department of Computer Science, Sindh Madrasatul Islam University,Aiwan-e-Tijarat Road, Karachi, Pakistan

Muhammad Jawed Iqbal Institute of Space and Planetary Astrophysics, University of Karachi, Karachi, Pakistan

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.

Publication Details

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.