Sunspot index infers a small relict magnetic field in the Sun's core

Abstract

Cowling1 long ago suggested the presence of a relict field in the Sun's core and argued that the bulk electrical conductivity was sufficient (provided that the field was dipolar) to prevent decay over the lifetime of the Solar System. Evidence for such a field is obtained by inference from the asymmetry in time of the solar sunspot number taken over many cycles2. As the Sun's dipole field reverses polarity approximately every 11 yr, if a core field of odd multipolarity (presumably dipolar) is present and in reasonable alignment with the oscillating dipole of the convective zone, it follows that in successive 11-yr periods, the two fields will tend first to be additive and then subtractive3,4. Because the net dipole field is the ‘seed’ field from which the much larger toroidal field arises5, the latter will vary in magnitude in alternating half cycles and correspondingly mediate the level of sunspot activity. Whether a fossil or relict field is present can be explored either by a search for the presence of a 22-yr line quintet in the Fourier spectrum of the sunspot index2, or alternatively the time sequence itself can be processed so as directly to bring out evidence for an offset. The latter involves a ‘demodulation’ of the sunspot sequence to emphasize the basic 22-yr Hale cycle. The latter is discussed here, but as the two are interrelated, the former is reviewed briefly first to provide the basis for the subsequent least squares (LS) analysis of the demodulated signal. As the solar cycle is unstable to drifts in frequency, the presence of frequency modulation can lead to a spurious relict field signal. A short summary of limits on frequency modulation (FM) contamination of the sunspot index spectrum is also given.