Statistical linguistic characterization of variability in observed and synthetic daily precipitation series

As an application of the proposed methodology, the same analysis was applied to stochastic time series obtained from standard weather generators. We found that the language defined by a first-order Markov process agrees with the observed language of the observed series.