The purpose of this work was to compare three methods for simulating data of photosynthetically active radiation and air temperature, using daily parameters of 17 stations of São Paulo State, Brazil. The simulation of those elements was carried out for de 1st and the 16th day of each month, through three cases: nonsymmetric triangular distribution; normal distribution truncated at 1.96 standard deviation; and bivaried normal distribution. The simulated data were evaluated through the tests of homogeneity of variance F and Bartlett, t test, agreement index of Willmott, angular coefficient of the straight line, the index of performance of Camargo and tack the normal distribution (uni-varied). The simulation using the bi-varied normal distribution is most appropriate for representing the climate variables.
probability distributions; climatic variables; bivariate normal; normal distribution; triangular distribution
