Markov Chain Application for Dry and Rainy Days

Abstract Rainfall is an important meteorological phenomenon in the tropical region, characterized by its spatial-temporal variability and associated extreme events. Thus, prior notification of the occurrence of one day being rainy or dry is extremely important for many human activities, especially for agriculture. The aim of this study was to analyze the occurrence of dry and rainy days in Rio Largo-Alagoas, Brazil, using the Markov chain. Daily rainfall data between 1973 and 2008 were used. It was considered six precipitation levels for dry and wet days and, it was applied to Markov chain to identify the probabilities of conditional occurrences of dry and rainy days. The study region had dry (September to March) and rainy (April to August) seasons well defined considered limits of precipitation values between 0 and 2 mm. Higher occurrences of dry and rainy days occurred from November to December (94%) and June to July (84%), respectively. The Markov chain concluded that the transition between dry and rainy days is low throughout the year.