The studies involving rainfall are quite difficult to conduct due to the fact that the available historical series, in most of the cases, are too small to make them. Thus, it is necessary to adjust a stochastic simulation technique that facilitates the generation of daily rainfall values, using a spreadsheet. In Piauí State, studies with this objective are inexistent for its several regions of agricultural importance. This work was executed with the objective of simulating the daily rainfall for the regions of Parnaíba and Teresina - PI, Brazil, using a spreadsheet. The simulation process was based on Monte Carlo's method. The occurrence of the rainfall was determined through a Markov chain, while its magnitude was estimated by gamma and empirical distribution functions. The model used provided estimates with excellent precision and accuracy, checked by satisfactory results of the statistical performance index.
Markov chain; Monte Carlo's method; gamma distribution
