Separating polar ionospheric and lithospheric magnetic signals in satellite data
Smith, Ashley Robert Andrew
Both the interior of the Earth and near-Earth space contain a variety of interacting magnetic field sources, which together make up the “geomagnetic field”. When a magnetic field measurement is made, either on Earth or in space, a superposition of these sources is observed. In order to study each source in isolation, it is therefore necessary to process these collected data to separate the superposed signals. Over the past two decades, the volume of data available has increased tremendously, in particular due to a number of satellite missions carrying high quality magnetometers. This has motivated the development of new techniques to make greater use of the available data. This thesis uses data from the contemporary European Space Agency mission, Swarm, to investigate sources primarily within the ionosphere and lithosphere, focusing on the polar regions where their mutual contamination is most extreme. I also use data from older missions: CHAMP, Magsat, and POGO. Of the ionospheric field sources, the auroral electrojets are one of the strongest and most variable. These are horizontal electric currents that flow within the auroral regions (at high latitudes). Their magnetic signal is not only an important aspect of auroral studies, but is also particularly difficult to separate from lithospheric field models derived from satellite data. I use a combination of satellite field intensity measurements from the past 50 years to examine their climatology, investigating their average behaviour in relation to solar wind, solar cycle, local time, and seasonal factors. I identify inter-hemispheric asymmetries, solar cycle modulation, and a secular shift in their location due to the variation of the core field. I also discover features which indicate incomplete removal of the lithospheric signal, showing that studies of the weakest ionospheric currents are still contaminated by lithospheric sources. This can be remedied by improved lithospheric models. In the latter part of the thesis, I create a lithospheric field model with a new coriii rection technique to reduce the effect of ionospheric noise. As part of this process, I demonstrate a bias field over the poles due to the noise introduced by the auroral oval. The technique involves stacking signals onto two global icosahedral hexagonal grids, one in geographic coordinates and one in local time coordinates, respectively organising lithospheric and ionospheric sources in their appropriate reference frames. The signals on the ionospheric grid are then removed from the input data and the reduced (stacked) data in geographic coordinates are inverted for a spherical harmonic model. This is demonstrated with Swarm data and shows a beneficial effect, but more work is needed to bring the model to the precision of the current state-of-the-art models. These two studies are directed toward two different physical systems but the observations are derived from the same data. In each case, attempts are made to remove the opposing “noise” from the target signal. In this way, the data are interrogated from both perspectives, where that which is considered as “noise” in one case is used as “signal” in the other.