IMPACT OF TROPICAL SEA SURFACE TEMPERATURE ON WEST AFRICAN MONSOON – A CASE STUDY OF SELECTED STATIONS IN NIGERIA

Abstract

Changes in tropical Sea Surface Temperature (SST) is one of the forces that drive the monsoon variabilities over West Africa. Therefore, this study investigates the teleconnection and covarying relationships between rainfall and SST in the tropics during the period from 1982 to 2013. Rainfall data covering nine (9) stations in Nigeria were obtained from the archives of Nigerian Meteorological Agency (NIMET), and SSTs were acquired from Optimum Interpolated Sea Surface Temperature (OISST) over eight ocean basins during the period of study. ERA Interim reanalysis zonal and meridional wind components and specific humidity at the surface, 850 hectopascals, 700 hectopascals, and 200 hectopascals were also obtained for the study period. The study adopts wavelet coherence analysis to perform spectral analysis of temporal variations and localized co-varying relationships between rainfall in the nine Nigerian stations and SST of eight ocean basins. Pearson correlation method was further used to determine the impact of SST on the tropical rainfall. Results show that SST in the ocean basins accounts for ~13% of the variations of rainfall in the stations. The power spectra revealed several localized regions with high powers at various periodicity bands, which implied that there are significances in the rainfall of the nine stations at various periodicities. The phase differences showed an antiphase (rainfall lagging) between the rainfall from the 9 stations and SSTs from the East Atlantic, West Atlantic and Indian Ocean and an in-phase (SST leading) between the rainfall from the 9 stations and the SST from the Mediterranean Sea significant at 95% confidence level. The correlation results showed that there were deficiencies in the representation of the connection between rainfall from the nine stations and SSTs of the Nino regions. However, the comparison of the correlations of SST in the ocean basins and rainfall in the 9 stations showed strong relationships.