The objective of this study was to analyze the spatial variability of the rainfall erosivity indices for the State of Rio de Janeiro, Brazil, using geostatistical analysis. Rainfall erosivity indices EI30 (defined by the product of kinetic energy of the rainfall and its maximum intensity on 30 minutes), and KE>25 (defined as the kinetic energy of the rain just to rainfall intensity higher than 25 mm h-1) were calculated using rain charts for 36 stations, while for the remaining 57 locations, this indices where estimated using standard regression procedures with rain gauges data. Therefore, the total number of points for the entire state was 93. The experimental semivariograms calculated for both indices were exponential. Using the model fitted to the experimental semivariograms it was possible to interpolate values for the locations where they were not known using kriging, which has advantages among other methods. Moreover, kriging variance maps were also generated. The largest values found for the rainfall erosivity indices were located at the Sierra regions and at Ilha Grande Bay, while the smallest values occurred at the Northern and Northwestern Regions of the state. The largest kriging variance values occurred at the Coastal and Northern Regions, where the data density is the smallest.
geostatistics; rainfall erosivity indices; kriging
