An Alternative Picture of the Geomagnetic Field.

Abstract

Spherical harmonic models of the Earth's magnetic field are a convenient way of describing the characteristics of the field. It has been common practice to use the Lowes/Mauersberger function to describe the power of the field in each degree of harmonic. There are several characteristics of this power spectrum which are well known from recent analyses of the geomagnetic field. Firstly, there is a general decrease of power as the degree of harmonic is increased. This general pattern is that the power varies as a function of a constant raised to the power n where n is the degree of the harmonic. However, there are some exceptions to this rule of thumb. Firstly, the dipole term (n = 1) is higher than expected. Secondly, the degree two term is lower than expected. Thirdly, the degree eight term is lower than expected. Although these facts might be considered to be just a 'chance happening, characteristic of today's field but not of other fields, it is shown that these characteristics are true for the last 200 years, during which time both the degree two and the degree eight individual harmonic terms change drastically. It is therefore suggested that these patterns represent a more permanent feature of the field, which may be caused by the sources approximating current loops located close to the core-mantle boundary (CMB). Another interesting fact about the spherical harmonic analyses of the field over the past 200 years is that the low order harmonics tend to be significantly stronger than the high order harmonics. This probably reflects sources which are preferentially located at higher latitudes, or are stronger at higher latitudes than at lower latitudes. The rates of change ofthe field are also studied, and it is shown that these rates of change are similar to what might be expected if the foci of the field at the CMB moved without changing their strength greatly. These simple observations based on analysis of the Gauss coefficients give information which may be useful for dynamo theories of Earth's magnetic field.

Secular Variation of the Earth's Magnetic Field.

“…Almost all of the loops have very large radii, mostly subtending half angles at Earth's center of 60° or more. The exception is loop 5 in models C5, C6 and C7, which has a half angle of 19° to 35°, roughly the size of the current loops discussed by Harrison (1994). 5.…”

“…4. It is probable that the sources of the non-dipole field should look like current loops located at the core mantle boundary (Harrison, 1994). In order to achieve a large variation of the angular standard deviation of VGPs as a function of observation latitude, it is almost certainly necessary that the frequency and/or strength of the sources should increase considerably as a function of absolute latitude.…”

“…5. Most of the change in the field is caused by the movement of the core surface sources and not by their waxing and waning (Harrison, 1994). 6.…”

“…The dipole and quadrupole components show a small increase, and the total field shows an increase from about 14.3° in 1690 to almost 22° in 1990. Although the reliability of the data for the very early part of this time interval is doubtful (Harrison, 1994), it seems likely that at least for the past 200 years, Earth's surface field is well represented by the spherical harmonic coefficients in these models of Bloxham and Jackson (1992). The field around 1770 should be especially well defined since there is a peak in the data availability at this time.…”

“…For instance, the Giant Gaussian model (Constable and Parker, 1988;Constable, 1990) is a model in which the spherical harmonic coefficients for the field analysed at the core-mantle boundary are chosen using a random number generator. However, it has been shown that within each degree of harmonic, the Gauss coefficients for the higher orders tend to be considerably smaller than those for lower orders (Harrison and Huang, 1990;Harrison, 1994), and the Gaint Gaussian model does not produce this effect. It has also been pointed out (Kono, 1993) that this distribution does not produce the latitudinal variation in VGP scatter seen in the present day field and in the PSV data.…”