| Dual frequency GPS receivers enable the estimation of absolute ionospheric delay and total electron content (TEC) along the signal path. By using a number of reference stations, each equipped with a dual frequency receiver, it is possible to estimate values of the vertical ionospheric delay at a set of designated grid points (in latitude and longitude) on an ionosphere shell. This type of ionosphere delay modelling is employed in wide area differential GPS (WADGPS) networks, where grid accuracies generally depend on the temporal/spatial correlations of TEC. These models can suffer degraded performance in regions, such as the high latitude auroral zone, where spatial gradients and temporal variations of electron density may differ significantly from assumptions. Given that future extensions of WADGPS applications include implementation of safety-critical systems to support air navigation in auroral regions, such as the wide area augmentation system (WAAS - FAA[1994]), it is important to establish the impact of auroral effects on such GPS networks.; In the research presented here, a detailed study of auroral effects on GPS is conducted, and performance of the ionosphere grid model is investigated in the auroral region. The development of a Canadian wide area network in 1996, by Natural Resources Canada (NRCan) [Caissy et al., 1996], allowed compilation of an extensive GPS data set from the auroral region. In this thesis, periods of enhanced auroral activity (over Canada) are identified from ground-based magnetometer signatures, using the Canadian Space Agency's CANOPUS MARIA array. Signatures of auroral TEC associated with these disturbances are established using corresponding GPS observations from the NRCan network. A wide area ionosphere grid model is developed, and degraded grid accuracies associated with the auroral features are estimated, under various modelling assumptions. It is observed that grid accuracies are degraded by a factor of 2–5, relative to typical accuracies, during periods of enhanced auroral activity. In order to mitigate these effects, a modified grid algorithm is developed in which auroral disturbances are detected from time series of TEC observations. Model parameters are adjusted, consistent with local ionospheric processes, which allows improvements in the grid performance of 15–25 percent. Detailed analysis of a substorm event is conducted, where the largest degradations in grid accuracies are observed during the substorm expansive phase. The potential of monitoring auroral phenomena using GPS observations from a wide area network is explored briefly. |
