Periodic signals in Spanish monthly precipitation data

Summary Forty-two monthly rainfall time series from Spain have been analyzed for the presence of cycles and regional patterns. The study was carried out by means of periodicity tests devised by Whittle, Hannan, Bartlett, Siegel, and Priestley. These tests contribute to a mixed spectrum analysis leading to the evaluation of signals versus the background of coloured noise (autocorrelation). Also, a direct maximum entropy method (MEM) for estimating spectra was applied to the series, in particular an autoregressive (AR) power spectral density (psd) estimation, and an autoregressive moving average (ARMA) psd estimation, with parameters obtained by exact maximum likelihood estimator (MLE).With 4 Figures