An experimental study of palaeosecular variation

Abstract

A total of 83 selected palaeomagnetic results are used in a latitude analysis of palaeosecular variation to show that Cox's 1970 model for secular variation fits the data, provided the dipole wobble is between 1 1/2° and 2°. The major contribution to secular variation then arises from the joint effect of dipole oscillations with the non-dipole terms. The parameters required to produce this effect are in accord with those used in Cox's 1968 model for reversal frequency. In time, secular variation for the pre-Tertiary is about 15 per cent smaller than during the Tertiary. An alternative approach using large equatorial collections confirms that palaeosecular variation must be due to a combination of dipole wobble, dipole oscillations, and non-dipole effects, with dipole wobble again making only a small contribution.