| Summary: | Magnetic surveys are an important method of understanding subsurface geology, however there are several reasons why correction by remote referencing may fail, including local induced effects, activity levels of the field, and simple distance between survey and base station. We look for ways to improve correction by remote referencing using insights from global field models and comparisons of data from a wide range of observatories. We investigate the conditions in which the behaviour of nearby observatories differ from each other, and where the CM4 comprehensive model fails to match the observed behaviour of the local geomagnetic field. The misfits are separated by cause: those due to the activity level of the geomagnetic field, and the location of the observatory. We see that CM4 is a good match to observatories in the conditions it was designed for (mid-latitudes and Kp up to 2), but also that it can produce a good fit to stations out of this range (up to Kp of 3 or 4). The correlation of misfits to CM4 allows us to separate effects due to latitude, and location on the coast. Further investigation allows us to suggest some corrections that may improve the quality and extent of magnetic data gained by surveys in these locations. High latitude stations show changes in behaviour which fall into latitudinally split groups, most likely due to the presence of induced fields from ionospheric currents. Ensuring base station and survey fall into the same grouping would eliminate many of the problems this causes. Geomagnetic storms often lead to survey data being unusable due to their effects. We find that while X component data contains mostly storm signal, the Y and Z components at many stations contain retrievable data. The recovery period of the storm can, for most stations, be used after a regression is applied. We also consider the effects of induced fields due to the tides and the coast effect -well-known effects that can be seen at many stations. We find a correction for the dominant M2 tidal effect using cosine waves. We also find an approximate correction for the coast effect, using cosine or sine waves of the Sq period as appropriate for the station pair chosen. It is also noted that small differences in location can have a large effect on the induced fields, as seen at GUI and TAM, where storms seem to have a smaller than expected effect.
|
|---|
