TRENDS AND CYCLES OF EXTREME RAINFALL EVENTS AND THE IMPLICATION FOR THE DESIGN OF HYDRAULIC STRUCTURES IN NIGERIA

Abstract

The human society has remained vulnerable to extreme rainfall events causing severe loss of human lives and valuables across many sectors. Global warming effects have continued to drive considerable temporal and spatial shifts in variability and change of Nigeria’s climate. Therefore there is a need for studying continuous change in characteristics of extreme events as related disasters are getting more frequent and severe. This research was as such aimed at identifying trend in extreme rainfall events and its accompanying effect on design of hydraulic structures in Nigeria. Extreme rainfall events were calculated as annual indices from a 41 year (1975 – 2015) daily rainfall data from fourteen stations across Nigeria. The frequency indicators were represented by number of rainy days where daily rainfall is more than or equal to 1 mm, 20 mm, 50 mm and 95th percentile while intensity indicators were represented by maximum daily rainfall, maximum 5-day rainfall, total annual rainfall and simple daily intensity index. The thresholds were defined as daily rainfall value at 95th percentile and for 1, 5 and 25 year return periods using ranked accumulated times series and Gumbel extreme value index method. Results showed average values of 51.5 mm, 83.8 mm, 107.3 mm and 145.6 mm for 95th percentile, 1, 5 and 25 year thresholds respectively. Quality control such as outlier and missing values checks and homogeneity test were applied to the data prior analysis. The climatological mean of indices was tested against elevation which produced a strong correlation and showed topography factors can on an average describe about 40% of the indices variation. Temporal trend of indices was tested using the non-parametric Mann-Kendal test while Sen’s slope was used in estimating trend slope. The trend analysis showed a spatial pattern characterized by high frequency and intensity indicator values in the rain forest agroecological zone, and low frequency and intensity in the savannah zone. The dominant trend discovered was the non-significant positive trend. However vi Jos consistently showed negative non-significant trend for almost all of the indices. Bauchi, Ilorin and Lokoja also showed significant positive trend in at least four of the indices and have been identified as high risk areas prone to severe extreme related disaster in the future. The identified dominant trend may render existing hydraulic structures incapable and inadequate to meet rising challenges in future and as such may require upward review of existing design criteria.