With the aim of verifying the temporal and spatial variability of the sample size ten-day average of the global solar radiation of 22 locations of Rio Grande do Sul State, Brazil, the global solar radiation data were collected from 1956 to 2003. One determined the sample size in each ten-day and location by clustering it using the complete linkage method. Results show that there is variability in the sample size of the ten-day average global solar radiation in Rio Grande do Sul State both in temporal and spatial scale. Greater sample size is necessary in the ten-day of June, July, August and September in relation to the others. For the studied locations and ten-day, 30 years of data are enough to predict the global solar radiation average (µ), for an error estimation equal to 12.3% average interval, with a reliable degree of 95%.
Sampling; spatial and temporal variability; inferential statistics
