The crop yield in a semi-arid region is highly dependent on the interactions between their phenological phases and the climatic variations of the area. In this study sea surface temperature (SST) anomalies of the Atlantic and Pacific oceans and annual yield of crops bean corn, cotton and cassava cultivated in the semi-arid region of Northeast Brazil were used, corresponding to the agricultural period from 1974 to 1996. Considering only the statistically significant anomalies of SST of the Atlantic and Pacific oceans in the period from November to March, models for forecasting yields of these crops in a semi-arid region as a function of the oceanic phenomena El Niño and Atlantic Dipolo were determined. The results suggested that such models could be used with reasonable accuracy in the identification of strategies for reduction of the impacts of the climatic variations on yields in the semi-arid region of Northeast Brazil.
SST anomalies; crop yield; forecast model
