A spherical harmonic analysis of solar daily variations in the years 1964-1965: response estimates and source fields for global induction-II. Results

Abstract

Inductive responses are derived for six harmonics of geomagnetic daily variations, ranging from 24 to 4 hr in period and from 650 to 350 km in penetration depth z*. Simultaneous data from up to 90 magnetic observatories are involved, and the final results are least-squares estimates with 123 quiet days in a new definition of magnetic quietness. They utilize the potential method of geomagnetic deep sounding, based on series expansions of the magnetic surface field into spherical harmonics, and thus lead to global response estimates in reference to the Earth as a whole. The analysis is carried out as a univariate linear regression, either between expansion coefficients for the vertical field component and the unseparated potential, or between internal and external potential coefficients, involving assemblages of single days or months. Attempts have been made to remove contamination by lunar daily variations and by internal induction anomalies. The analysis yields response estimates for all spherical harmonic terms in the series, and in the final analysis 12 terms are used for each time harmonic. Estimates for four local-time terms are similar, while those for eight general terms are scattered for the first harmonic, but converge towards higher harmonics. Anomalous effects, as predominantly expected from induction in the oceans, seem to concentrate on low-order time and spherical harmonics, and some degree of lateral uniformity of conductivity apparently exists within the depth range into which daily variations penetrate. This conclusion may be biased towards continents in the northern hemisphere, where most observatories of this study are situated. The results obtained are nearly the same, whether single days or monthly averages are analysed and whether all days or only quiet days are used. The derived responses for the principal terms of the six time harmonics are compatible with local estimates from other methods, averaged over continental sites, but they are more precise. They agree with responses of a three-layered earth model from a previous joint interpretation of combined Sq and Dst results. Problems arise with the fifth and sixth time harmonics, which in their spatial structure seem to be detached from the global pattern of daily variations. Expansion coefficients for the local-time terms on quiet days show a distinct seasonal variability in the form of annual and semi-annual variations, for both their external and their internal parts, which are closely correlated on a day-by-day basis. External coefficients, representative of the four seasons and with corrections for lunar daily variations, are tabulated. Their equivalent currents are displayed on global maps. Special attention is paid herein to the zero reference level in the hypothetical absence of any transient variation field.