A Seasonal and Heteroscedastic Gamma Model for Hydrological Time Series: A Bayesian Approach

Time series models are often used in hydrology to model streamflow series in order to forecast and generate synthetic series which are inputs for the analysis of complex water resources systems. In this paper, we introduce a new modeling approach for hydrologic time series assuming a gamma distribution for the data, where both the mean and conditional variance are being modeled. Bayesian methods using standard Markov Chain Monte Carlo Methods (MCMC) and a simulation algorithm introduced by [1] are used to simulate samples of the joint posterior distribution of interest. An example is given with a time series of monthly averages of natural streamflows, measured from 1931 to 2010 in Furnas hydroelectric dam, in southeastern Brazil.