EPIC (Erosion Productivity Impact Calculator) model has been developed to simulate the effect of agricultural management strategies of water resources and soil productivity. In this study, the performance of meteorological data simulators were evaluated for six (Campinas, Jaú, Manduri, Mococa, Pindorama and Ubatuba) localities in the State of São Paulo, Brazil, which include the Markov chain probabilities for the sequences of wet and dry days; the maximum air temperature model adjusted for wet days; and the solar radiation model also adjusted for wet days. From tests and adjustments of the models using the historical data for each locality, the best performance for the model of probabilities for dry days sequence has been found, taking into account the last wet day - P(W|D) in relation to the model of probabilities of two consecutive wet days sequence - P(W|W). The maximum air temperature and solar radiation models have presented good performance in tests after the adjustment for each locality. For all models, the best results have been obtained from the local adjustments, when compared to the original model conceived for EPIC. Model results for Ubatuba data were distinct from the other sites.
climate; erosion models; probability
