Evidence for geomagnetic jerks from 1931 to 1971

Abstract

Examination of annual mean values of the geomagnetic elements from magnetic observatories worldwide has shown that a sudden change in secular acceleration, which has been termed a jerk, took place at about 1970. Malin et al. used the set of secular variation models compiled by Hodder to provide a global description of the jerk, and attempted to quantify it by computing the mean square value of the jerk field over the Earth's surface. To determine the ability of the method of Malin et al. to distinguish the 1970 jerk from noise their analysis has been repeated at 2-y intervals from 1931.5 to 1971.5, a period during which no other similar jerk is thought to have occurred. It is shown that the mean square value of the jerk field is not a reliable indicator of the existence of a jerk when taken alone as noise contributes to it additively giving rise to “apparent” jerks. This point is illustrated by examples of contour maps of the jerk field for 1949.5 which show how large mean square values relate to spurious foci in regions where there are no magnetic field observations. These maps contrast with those for the 1970 jerk which relate well to the observations and it is shown that the 1970 jerk is well-resolved from noise and is unique within the time-span considered.