| Affiliation: | 1.Departamento de Medicina Social, FMRP,Universidade de S?o Paulo,Ribeir?o Preto,Brazil;2.Departamento de Matemática,Universidade Tecnológica Federal do Paraná,Cornélio Procópio,Brazil |
| Abstract: | Rain precipitation in the last years has been very atypical in different regions of the world, possibly, due to climate changes. We analyze Standard Precipitation Index (SPI) measures (1, 3, 6 and 12-month timescales) for a large city in Brazil: Campinas located in the southeast region of Brazil, São Paulo State, ranging from January 01, 1947 to May 01, 2011. A Bayesian analysis of non-homogeneous Poisson processes in presence or not of change-points is developed using Markov Chain Monte Carlo methods in the data analysis. We consider a special class of models: the power law process. We also discuss some discrimination methods for the choice of the better model to be used for the rain precipitation data. |
