ANÁLISE DA ADERÊNCIA DE DISTRIBUIÇÕES DE PROBABILIDADE AOS DADOS DE TEMPERATURA MÁXIMA E MÍNIMA DO AR EM IGUATU-CE

Abstract

This study was aimed to adjust probability distributions to a data series of maximum and minimum daily air temperature of Iguatu city in different periods of time, for the purpose of observe that the distributions under study has a better performance and has recommended its use to estimate the probability of occurrence of the estimated values of maximum and minimum air temperature. For the adjustment the data in distribution of frequency were used six probability distributions: Beta, Gamma, I Gumbel, Log-Normal, Normal, and Weibull. Through the adherence tests of Kolmogorov-Smirnov at 20% of significance and Chi-square at 5%, the distributions Normal Log-normal and have a better fit to the data for all scales analyzed, recommended the use of normal function by ease the estimation of its parameters and probabilities.