The objective of this study was to establish sequences of rainy days as a risk factor for agriculture in the southeast region of the states of Goias, Brazil. Rain gauge data from 36 rain stations were used. The sequences of rainy days were calculated by a computational routine for December, January, February and March. The probability of sequences was estimated by using the TNBD (Truncated Negative Binomial Distribution) and was verified by Chi-square distribution. The result were regionalized obtaining figures related to sequences of rainy days for the months studied. Results showed that the risk of rain sequences varies locally. Higher probability for occurrence of two rainy days occurred in the center, southwest and south of the analyzed region, and 3 days in the central region and surroundings. There was low probability for occurrence of 4 or more rainy days.
binomial distribution; sequences of rainfall; rain gauge data
