RAINFALL ANOMALY PATTERNS IN DRY AND WET YEARS OVER NIGERIA

Rainfall amount and rainfall frequency for five individual dry and five individual wet years were compared with the 1941-1987 averages for Nigeria. Five spatial-anomaly types were found to emerge for dry years, namely above average rainfall in southern Nigeria but below average rainfall in northern Nigeria (spatial type I), above average rainfall in the centre but below average rainfall in the coastal and extreme northern parts of the country (spatial type II) and below average rainfall in southern Nigeria but above average rainfall in northern Nigeria (spatial type III). The other two anomaly types are country-wide occurrence of below average rainfall (spatial type IV) and country-wide occurrence of above average rainfall (spatial type V). These spatial-anomaly types also hold true for wet years, except that the spatial type II is replaced by the occurrence of below average rainfall in the centre but above average rainfall in the coastal and extreme northern parts of the country (spatial type VI). The spatial type I and II anomalies support the hypothesis of a restricted northward advance of the Inter Tropical Discontinuity (ITD) in dry years and on a complementary basis the spatial patterns for the rainfall amount anomaly (RAA) and the rainfall frequency anomaly (RFA) depicted these two anomaly types in 40 per cent of the monthly periods for the set of dry years considered. The spatial type IV anomaly, which occurs during the peak of the rainy season, supports the hypothesis of a weakening of the rainy season intensity and on a complementary basis the spatial patterns for the RAA and the RFA depicted this pattern in 30 per cent of the monthly periods for the dry years studied. This shows that both hypotheses are valid for explaining dry years in subtropical West Africa. The spatial type III and V anomalies support the hypothesis of a considerable northward incursion of the ITD in wet years. It was found that the spatial patterns of the RAA depicted these anomaly types in 53.3 per cent of the monthly periods whilst the spatial patterns of the RFA depicted the anomaly types in 66.7 per cent of the monthly periods for the set of wet years considered.