Effects of long-term variation on the frequency spectrum of the geomagnetic reversal record

Abstract

Several studies have suggested a period of about 30 Myr in the frequency of geomagnetic reversal1–4 superposed on a long-term variation, but none of these studies explicitly investigated the effects of the long-term variation on their analyses. We show that the long-term variations alone account for most of the spectral evidence taken to support the 30 Myr periodicity. In particular, high peaks in the spectra near 30 Myr, the overall pattern of peaks, and shifts of the peak pattern for subsets of the data are explained well by a long-term variation model. We suggest that there may be periodic changes in reversal frequency about the long-term variation, but that more study will be required before the periodicity can be established.