Search

Geomagnetic induction vectors were calculated at the IMAGE magnetometer stations in Fennoscandia using hundreds of four-hour events of quiet time variations, and of magnetic storms due to eastward and westward electrojets. Quiet time vectors were found to be very similar to those calculated previously using a set of carefully selected plane wave events. Due to the source effect during electrojet events, the average direction of the induction vectors may rotate tens of degrees compared to the quiet time. The length generally increases tens of percent, but at the stations MAS and SOR near the electrojet and highly-conducting anomalies, a decrease is observed. If the quiet time induction vector is longer than about 0.50 then the source effect does not affect the direction of the vector remarkably even if the length may change considerably. If the conductivity anomaly is weaker (quiet time vector shorter than about 0.30) then the source effect dominates near the electrojet (PEL, MUO, KEV, KIL), but is less important farther away (NUR, OUJ). If a large number of magnetically quiet events are available of local morning hours, no special reduction of the source effect is needed in the research area.

The usual plane wave assumption of a sufficiently uniform primary field is questionable at auroral and equatorial latitudes near localized ionospheric currents. Besides attempting to select only events with distant sources, it is reasonable to try to eliminate the distortion due to source effects by taking averages over several events. However, it is difficult to determine the best averaging method in the calculation of the apparent resistivity, the impedance phase, or the induction vector. In this paper, we use synthetic models to demonstrate the differences in averaging methods.

A method to estimate geomagnetically-induced currents (GICs) in a power grid based directly on magnetic recordings is presented. To determine the electric field on the earth's surface for the calculation of GICs based on available magnetic data, a model is used in which the geomagnetic variation field propagates as a vertical plane wave in the earth and the earth is a half-space with a constant conductivity. To get statistically reliable results, the earth's conductivity is allowed to be a function of the geomagnetic K-index. The reason for this unphysical behaviour, which means that the conductivity is, of course, not real but some kind of an effective conductivity, is explained.