Rainfall erosivity at Teodoro Sampaio was studied through methodology proposed by Wischmeier and Smith (1958). A set of 812 erosive individual rains was selected, from a continuum series of nineteen years of nonrecording rain gauge data, according to Wischimeier (1959), Wischmeier and Smith (1978) and Cabeda (1976) and proposed with changes by Carvalho (1987). All the selected rains were computed by the package of Cataneo et al. (1982). The rainfall erosivity factor computed was 7172 MJ mm ha-1 h-1 y-1, which is expected to occur at least once each 2.33 years, with a probability of occurrence of 42.9 %. The values of the annual erosivity indices at that site expected in the return periods of 2, 5, 10, 20, 50, and 100 years were, respectively, 6,831, 8,666, 9,877, 11,046, 12,546, and 13,675 MJ mm ha-1 h-1 y-1. The greatest annual soil loss is expected to occur between the months of October and March, as shown by the 74.0 % erosivity 45.7 % with occurring between December and February of the total annual erosivity for that period. Therefore, aiming to follow the correct soil conservation principles, the better soil conservation pratices are recommended to be established and maintained throughout that particular period of the year for an effective erosion soil loss control. A high correlation was found between the montly average erosivity index EI30 and the rainfall coefficient at that site. Therefore, the regression equation obtained allows a fair estimate of the R factor to be used for other regions where climatic conditions are similar to Teodoro Sampaio's but without available data.
return period; probability of occurrence; rainfall coefficient; R factor; erosivity factor