Predicting The Variability And The Severity Of The “Little Dry Season” In Southwestern Nigeria

Abstract

This study evaluates the variability and severity of the ‘Little Dry Season' (LDS) in southwestern Nigeria. Being an important climatic phenomenon that impacts agricultural practices in the West African sub-region, a fore knowledge of the LDS characteristics for any given year should enable farmers to plan with greater confidence, and mitigate the negative consequences and exploit its beneficial opportunities. The study therefore examines the relationship between the LDS rainfall characteristics of southwestern Nigeria and the sea surface temperatures (SSTs) of the Gulf of Guinea, and of the source locations of the Guinea and Benguela currents, and generates models for the prediction of the LDS characteristics. Rainfall data of Ikeja, Benin City, Ibadan and Ilorin in southwestern Nigeria, are used in the analysis.The data were sourced from the archives of the Nigerian Meteorological Service, Oshodi, Lagos. The two sets of data are for a 38-year period (1961-1998). The data of 10 of the 38 years (1961-1970) were used for testing the model, while the remaining data (1971-1998) were used for building the prediction model. Linear regression algorithm and K-means cluster analyses were employed to generate the models. Results show that LDS rainfall amount and rainy days have significant positive relationship with SSTs of the Gulf Guinea and the source locations of the Guinea and Benguela currents. Very Cold, Cold, Average, Warm and Very Warm SST conditions promote Very Dry, Dry, Average, Wet and Very Wet LDS rainfall conditions, respectively. The time lag between the SSTs of the source locations of the Guinea and Benguela currents and SST of the Gulf of Guinea and thus LDS rainfalls, are 1-2 and 6-7 months, respectively. There is no time lag between the SST of the Gulf of Guinea and the LDS rainfalls. All the correlation values obtained are statistically significant. However, the ‘goodness of fit' assessment (by comparing the observed with the forecast) indicated that the model generated, using the SST of the source locations of the Guinea current, has a poorer ‘fit' than the model based on the source locations of the SST of the Benguela current. Given that both the statistical and real time tests of ‘goodness of fit' indicate the SST of the Benguela current origin as having the most coherent relationship, the study concludes that, the Benguela current is having preponderant influence on the dynamics of the Gulf of Guinea, and thus the LDS rainfall in southwestern Nigeria. Thus the SST of the Gulf of Guinea and the LDS rainfall in southwestern Nigeria can be reliably predicted with 6-7 months lead period, using the SST of the Benguela current origin. Keywords: Southwestern Nigeria; Gulf of Guinea; Benguela current; Guinea current; “Little Dry Season” (LDS); Sea Surface Temperature (SST); LDS rainfall amount; LDS rainy days IFE Journal of Science Vol. 9 (1) 2007 pp. 93-108