Dependence of inferred magnetic sector structure upon geomagnetic and solar activity

Abstract

Inferred solar sector polarity given by the AC index of Svalgaard, has been intensively studied as a single time series and as a time series correlated with geomagnetic and solar activity. Power auto-spectra of the AC index yield a highly significant harmonic series with fundamental at 27 days period and possessing clear harmonics up to the sixth; and a very prominent peak at a period of 1 yr. The 27 day harmonic series clearly indicates the solar control of the index while the 1-yr period might be taken as confirmation of the work of Rosenberg and Coleman to the effect that the sector pattern observed on Earth depends upon Earth's heliographic latitude which has a 1-yr period. Cross correlation analysis and superposed epoch analysis are used to show that sectors inferred to be positive or away are associated with low geomagnetic and solar activity whereas sectors inferred to be negative or toward exhibit significantly enhanced geomagnetic and solar activity. These results appear to be in conflict with superposed epoch analyses by Wilcox and Ness using satellite observed sector polarities which showed that geomagnetic activity increased after passage of a sector boundary, independent of the nature, whether + − or − + of the boundary. The conflict is resolved here by noting that the yearly correlation coefficient, at zero time lag, between inferred sector structure and geomagnetic activity averaged about 0·5 for the year 1927–1958, dropped to low values by 1960, recovered by 1962 and then dropped sharply in 1963 by an order to magnitude; the correlation has remained essentially zero ever since. Thus, the satellite results, all obtained post 1963, would not show increased activity during either sector sign. The results cast doubt upon the accuracy of the early ‘inferred’ sector polarities because it is felt that the only simple explanation for the strange behavior of the correlation coefficient lies in some artifact of the data.