ABSTRACT
This study aimed to obtain matrices of transition probability and the conditional monthly rainfall for the State of Santa Catarina, as an aid to climate forecast. Were used daily rainfall data for the period 1991-2013 from nine stations located in different climatic regions of the State of Santa Catarina. For matrices of probabilities, monthly rainfalls were divided in the following categories: below (-1), inside (0) and above (1) climatology. The framework of the values of monthly rainfall in these categories was based on quantile Q0,35 and Q0,65. When the rainfall is below the climatology (category -1) there is a greater probability of rainfall in the next month pass to category 1 in six regions, represented by the cities of Vargem Bonita (40%), Saudades (47%), Curitibanos (36%), Lages (40%), Trombudo Central (40%) and Araranguá (45%). A similar situation occurs when in a given month the observed rainfall is above the climatology, so there's more likely to remain in condition 1. The probability values show that the matrices of transition probabilities and conditional can be used as an aid in monthly forecasts and daily extreme events of rainfall to Santa Catarina.
Key words:
climatology; Markov chain; extreme events
Thumbnail
