Re-evaluation of possible historical relationship between magnetic intensity and climate

Abstract

HYPOTHESES concerning changes in the internal component of the geomagnetic field and consequent effects on climate1–10 are usually based on palaeomagnetic and palaeoclimatic data. Results are often ambiguous due to poor temporal and spatial resolution and difficulties in interpreting the proxy data10–12. Use of historic data partially avoids these problems. Thus, Wollin et al.13 tentatively conclude “that the trends in intensity from most of the magnetic observatories in the world with records over at least 30 yr correlate negatively with the 10-yr means of air temperature”. Because temperature and magnetic field intensity both have regional characteristics, this conclusion implies that the relationship is a local phenomenon although it will be valid globally. This further suggests that the relationship is causal. Wollin et al. presented graphs of mean annual magnetic intensities from 16 magnetic observatories along with 10-yr mean temperatures from nearby meteorological stations; one example is given using a 10-yr running average of winter temperatures and five examples use annual temperatures. Although the graphical data are suggestive of the proposed relationship, the study suffers in three respects. First, data from only 22 of the 43 magnetic observatories listed are presented. Second, the use of 10-yr temperature means yields only three or four data points which may not properly represent the trend of the temperature data. Third, the degree of correlation is not quantitatively evaluated. Here we present statistical evidence that there is no globally valid inverse relationship between geomagnetic intensity and climate for the period covered by observatory records.