The objective of this research was to estimate ten-day maximum (Tx) and mean (Tm) air temperature using altitude and the geographic coordinates latitude and longitude for the Rio Grande do Sul State, Brazil. Normal ten-day maximum and mean air temperature of 41 counties in the State of Rio Grande do Sul, from 1945 to 1974 were used. Correlation analysis and parameters estimate of multiple linear regression equations were performed using Tx and Tm as dependent variable and altitude, latitude and longitude as independent variables, for the 36 ten-day periods of the year. Pearson's linear correlation coefficient between estimated and observed Tx and Tm, calculated for tem counties using data of were used as independent data sets. The ten-day maximum and mean air temperature may be estimated from the altitude and the geographic coordinates latitude and longitude in the State of Rio Grande do Sul.
modeling; mathematics matrix; multiple linear regression; analysis of correlation
