Rainfall daily data of 600 meteorological stations were analyzed to assess the changes in precipitation in northeastern Brazil. Maps of rainy day numbers and the coefficient of variation (CV) for rainfall and rainy days are presented. Autocorrelation function was obtained for the annual rainfall in northeastern Brazil. Results showed that the semiarid regions have a high coefficient of variation for both rainfall and rainy days than those located in northeastern coast and ´agreste´ of Brazil. The highest values of coefficient of variation are associated to the lowest values of rainfall and rainy days. The rainfall change in northeastern Brazil is smaller in rainy regions than in dry regions. The high changes in rainfall and rainy days are limiting features to the rainfed agriculture in semiarid environments. The correlograms of rainfall indicated significant trend for the limits of 95% confidence levels on some stations.
correlograms; coefficient of variation; semi arid
