Geomagnetic data on variations of solar radiation: Part I—Wave‐radiation

Abstract

Summary—From geomagnetic observations, the time‐variations of two kinds of solar radiations can be deduced—a wave‐radiation, W, and a particle‐radiation, P. This paper, one of a series, derives and discusses homogeneous time‐series for W and P; these data, in addition to their meaning for geomagnetism and solar physics, may serve as numerical basis for studies on other solar influences in geophysical or biological phenomena. Daily values δW2 for the deviations of W from a normal value are inferred from suitably defined ranges of the solar diurnal‐magnetic variation of the horizontal intensity at the Huancayo (Peru) Observatory of the Carnegie Institution of Washington, for March, 1922, to December, 1937. Averages for eighths and for three‐eighths (= about ten days) of solar rotations are computed, also “smoothed decade‐deviations” showing the quasi‐persistent periodicities expressing the 27‐day recurrence‐tendency due to solar rotation. Monthly averages δW1 are extended to include October, 1939. Comparable tables are derived for solar activity R (from Zürich relative sunspot‐numbers R) and for particle‐radiation P (from data for geomagnetic disturbance). The correlations between R and W in “slow” variations (expressed in monthly, quarterly, and annual means) are the closest found so far between solar and terrestrial phenomena, surpassing even those found between R and P.The influence, on W and P, of changes in R in the course of solar rotations (“fast” variations) is studied by methods of correlation and by the superposed‐epoch method. Systematic features affecting the results of both methods are demonstrated; the relative exaggeration of the main selected pulse is recognized, explained, and illustrated in a statistical model. Several statistical experiments agree that—except near sunspot‐minimum—the fast variations of R are accompanied by similar variations of W, lagging by not more than about one day. Again, the statistical relation between R and P in the fast variations is found much weaker than that between R and W. Quantitatively, the relative effect of R on W in the fast variations is computed to be about 30 per cent smaller than in the slow variations, but reasons are given which interpret this result as compatible with the view that the relation between R and W in slow and fast variations does not differ essentially. The 27‐day recurrence‐tendency in W is just as strong as in R; there is an indication that the effect of a spot‐group on W increases with its age, if equal sunspot‐numbers are compared. The physical meaning of W is discussed; W is probably a solar radiation absorbed rather low in the ionosphere, in or near the same layer which is ionized by the excessive ultra‐violet emitted by a solar eruption. A program for the systematic extraction of W from geomagnetic records is outlined.Remark on symbols—Instead of using bold‐face sans serif type, as the practice adopted in the book “Geomagnetism” [see 3 of “References” at end of paper], light sans serif type has been used to designate, in general, the three phenomena R = solar activity as seen on the Sun's surface, W = wave‐radiation, and P = particle‐radiation effective in geomagnetism. The letter S was already adopted for solar‐diurnal magnetic variation, so R was chosen to represent Zürich relative sunspot‐numbers R. The letter C, which might have been suggested for corpuscular radiation, has been long used for daily international magnetic character‐figures (here Cint), so that the letter P for the effect of solar particles entering mainly in the zones of polar aurora was preferred. The indices in the deviations δW1 and δW2 indicate conventionally that these quantities have been derived successively. § 13 collects data on scales.