Short distance spatial variability for rainfall data was studied for Pindorama, State of São Paulo, Brazil, using 32-years-long daily rainfall data. Thirty-two-years-long daily rainfall series for 5 rain gauges, for which the separation distances range from 257 to 3900 meters, where submitted to correlation analysis for the assessment of the spatial variability. It was concluded that as the distance between the rain gauges increases the correlation coefficient between them decreases, which indicates that the rainfall depends on the spatial location of the rain gauges. When daily data was analyzed, it was found that even for rain gauges located close together but with difference in altitude, the differences in rainfall between them is higher than when they are farther apart but in similar altitude. With respect to the length of the series, it was concluded that they should have at least seven years of data.
spatial dependence; time series; correlation analysis
