Variation with a period of two solar cycles in the cosmic-ray diurnal anisotropy and the superposed variations correlated with magnetic activity

Annual means of the diurnal anisotropy from 1937 to 1967 are shown to result from the addition of two distinct diurnal components. One component, with maximum in the asymptotic direction 128° east of the sun, contains a well-determined wave, W, with a period of two solar cycles. W passes through zero in 1958 when the sun’s poloidal field reversed. The remaining component with W eliminated, has its maximum in the asymptotic direction 90° east of the sun. Annual means of this component, with maximum at 18.0 hours local asymptotic time, are well correlated (r = +0.75) with magnetic activity and determine a solar cycle variation with minimum near sunspot minimum and amplitude about two-thirds that of W. These results derive from a statistical investigation of the variability of annual means of the diurnal variation from ion-chamber data at Cheltenham-Fredericksburg, Huancayo, and Christchurch. The absolute, or total, diurnal anisotropy and the atmospheric diurnal temperature effect are in reasonable agreement with those derived independently through a comparison between the diurnal anisotropy from ion-chamber data and from Simpson's 1953–1966, IGY neutron monitor data at Huancayo.