Prediction of SPI Drought Class Transitions Using Markov Chains

摘要: Using the SPI relative to 67 years data sets, a Markov chains approach has been utilized for several locations in Alentejo, southern Portugal, characterize stochasticity of droughts, which allowed predicting transition from class severity another up 3 months ahead. models were applied using both homogeneous and non-homogeneous formulations. The results application are presented discussed, showing particular usefulness adopting formulation, allows differentiate predictions relation initial month considered, thus understanding probable evolution drought as influenced by climate and, particular, seasonality rainfall. However, these results, promising view management, require further developments be associated with other predictive tools stochastic or physical nature. Possible approaches on transitions risk management also discussed.