Statistical issues about solar-climate relations

Assessing the relationship between temperature variations and solar activity requires much physical and statistical insight. This paper is devoted to the latter. We focus on the statistical significance of diagnostics to obtain properties of the variability of time series. We illustrate our study by analyses of European temperature datasets and geomagnetic field variations. The goal of the paper is to provide a framework to control the spurious results that statistical tools can generate. We show that some variability diagnostics barely distinguish observed temperatures from auto-regressive random processes. In general, the variability diagnostics between temperature and geomagnetic activity are not significantly correlated, due to a low number of degrees of freedom.