Variations in the geomagnetic dipole 1: The past 50,000 years.

Abstract

An analysis has been made of all archeomagnetic intensity data for the past 50, 000 years. There are 1, 175 results from different parts of the world covering the past 12, 000 years but these are heavily biased to the region of the northern hemisphere between 0° and 90°E (64% of data). The global means show a broad maximum about 6, 500 years B. P. There is no evidence from data older than 7, 000 years that a quasi-cyclic variation with period about 104 years exists.A careful analysis of 472 dipole moments covering the 2, 000 year period between 1000 A. D. and 1000 B. C. shows that they have a normal distribution with standard deviation of 19.7%. The estimate of the standard deviation of results within each 1, 000-year-interval for the past 10, 000 years is 21.2%. Analysis of the present field suggests a standard deviation of dipole moments of 17.5% would apply if the present field is representative of the past. This suggests the standard deviation due to experimental errors in archeomagnetic results is about 10%, a value that concurs with independent analysis. The ten 1, 000-year-mean dipole moments have mean 8.75×1022Am2 and estimated standard deviation 18.0% which may be attributed to dipole intensity fluctuations.Over the period 15, 000 to 50, 000 years B. P. the earth's dipole moment was low, the 14 results available having mean value 4.44×1022Am2 with estimated standard deviation 26.5%. This is precisely the scatter to be expected from a combination of dipole intensity fluctuations of 18.0% non-dipole variations of 17.5% and experimental errors of 10% as deduced for the past 10, 000 years. The conclusion therefore is that variations in the non-dipole field always remain in the same proportion to the dipole field irrespective of its magnitude. Furthermore, the time scale of changes in the earth's dipole moment must be very much longer than has previously been supposed and must be at least 105 years.