OPTIMIZATION TECHNIQUE IN RAINFALL – INTENSITY – DURATION – FREQUENCY MODELLING FOR IKEJA, SOUTH-WEST, NIGERIA

Authors

  • AO David DEPT OF CIVIL ENGINEERING, OLABISI ONABANJO UNIVERSITY, IBOGUN CAMPUS, OGUN STATE, NIGERIA.
  • IL Nwaogazie DEPT OF CIVIL & ENVIRONMENTAL ENGR’G, UNIVERSITY OF PORT HARCOURT, RIVERS STATE, NIGERIA.

Keywords:

IDF models, Gumbel Extreme Value Type - 1, Normal, Log Pearson Type - 3 distributions, Excel Optimization Solver, goodness of fit test, Ikeja.

Abstract

The adequate estimation of rainfall intensity over a particular catchment is a necessary procedure in the design of water resources engineering control structures. To develop the probability and non – probability distribution function models for rainfall intensity – duration – frequency for Ikeja, 25 year daily rainfall data were collected from Nigerian Meteorological Agency (NIMET) Abuja for Ikeja. The annual maximum rainfall amounts with durations of 5, 10, 15, 20, 30, 45, 60, 90, 120, 180, 240, 300 and 420 minutes were extracted and subjected to frequency analysis using the Excel Optimization Solver wizard. To develop the rainfall intensity, duration and frequency (IDF) models, specific and general IDF models were obtained for return periods of 2, 5, 10, 25, 50 and 100 years using the Gumbel Extreme Value Type - 1, Normal and Log Pearson Type - 3 distributions. The Anderson Darling goodness of fit test was used to ascertain the best fit probability distribution. The R2 values ranged from 0.992 – 0.993 and the Mean Squared Error, MSE from 26.43 – 115.94 for the Gumbel; 0.992 – 0.993 with MSE of 28.64 – 85.23 for Normal distribution and 0.991 – 0.993 with MSE of 28.24 – 154.85 for Log Pearson Type – 3.The prediction of rainfall intensity with the Probability Distribution Functions showed a good match with observed intensity values. The intensity – duration curve in all cases has a negative slope. The GEVT – 1 and Normal distribution models ranked first while Log-Pearson Type 3 ranked third with respect to R2 and MSE in the non-specified return period. The probability distribution models are recommended for the prediction of rainfall intensities for Ikeja metropolis.

 

http://dx.doi.org/10.4314/njt.v39i4.37

Published

2020-09-30

Issue

Section

Agricultural, Bioresources, Biomedical, Food, Environmental & Water Resources Engineering