The aim of this study is to test whether it is possible to predict a Grand Minimum, assuming that there is, in the solar activity, precursor information about the approaching minimum. The monthly averages of sunspot numbers, covering the period from the year 1610 until the present, are used as input data. The time series is converted into a multivariate time series of indicators (the Multiple Indicator Model technique). The multivariate time series for periods including the Maunder and Dalton Minima is used to train a Neural Network model, which is later applied to recent solar sunspot data. The result shows a clear similarity between the periods before the Maunder and Dalton Minima and the period after the year 2000.
IRF Scientific Report 299
May 2009.