The potential of the Earth'S magnetic secular variation

Abstract

Data on secular variation given by Fisk for the epoch 1920–1925 are described. From these data are obtained new isoporic charts showing the geographical distribution of the north and east components of the variation. Tentative measures are applied to test the accuracy of the data. In making these tests it is assumed that the secular variation of the Earth's non‐potential field, if the latter exists, should be very small or zero, in the case of accurate magnetic data. Line‐integrals were evaluated for the horizontal component of secular change taken along parallels of latitude and meridians of longitude, and also around tesseral areas. The results indicated that these line‐integrals would vanish if the error in the data averaged along the path of integration were less than about 20 per cent of the observed secular change as given by Fisk, except in south polar regions where the average error along parallels of latitude is estimated to exceed 100 per cent. The values of the line‐integrals were found to vary in a systematic way over the Earth, thus strongly suggesting the presence of important systematic as well as accidental errors in world‐wide maps of the Earth's field.