Abstract. This paper presents a novel method to determine a baseline for magnetometer data. This baseline consists of all magnetic field components not related to ionospheric and magnetospheric disturbances, i.e. all field components due to solar quiet variations and other background variations, such as tidal and secular variations, as well as equipment effects. Extraction of this baseline is useful when the magnetic field variations due to solar disturbances are analysed. This makes magnetometer data suitable, for instance, for the calculation of ionospheric equivalent currents related to geomagnetic storms and substorms.

The full baseline is largely composed of two main constituents: the diurnal baseline and the long-term baseline. For the diurnal baseline, first "templates" are derived, based on the lowest few harmonics of the daily curves from the quietest days. The diurnal variation of the baseline is obtained by linear interpolation between these templates; this method ensures a smooth baseline at all times, avoiding any discontinuities at transitions between days. The long-term baseline is obtained by linear interpolation between the daily median values of the data; this way the baseline is ensured to follow long-term trends, such as seasonal and tidal variations, as well as equipment drift. The daily median values are calculated for all days except the most disturbed ones; a procedure for this selection is included.

The method avoids many problems associated with traditional baseline methods and some of the other recently published methods, and is simpler in procedure than most other recent ones. As far as can be compared, the distribution of the resulting field after removal of the baseline is largely similar to that using other recent baseline methods. However, the main advantage of the method of this paper over others is that it removes equipment drift and other artefacts efficiently without discarding too much data, so that even low-quality data from remote unmanned magnetometers can be made suitable for analysis. This can give valuable contributions to the database of ionospheric equivalent currents, especially in the area near the polar cap boundary.