Statistical Analysis of GPS Ionospheric Scintillation and Short-Time TEC Variations over Northern Europe

摘要

With the increasing demand for real-time high-accuracy positioning services, possible sources of GPS errors have been intensively investigated and solutions to minimise their effects have been formulated. Within this context, the ionosphere represents one of the most important sources of accuracy degradation. The ionospheric total electron content (TEC) is responsible for range errors due to its time delay effect over transionospheric signals. Also, electron density irregularities in the ionosphere are responsible for fluctuations in the phase and amplitude (scintillation) on the GPS signals. During amplitude scintillation, the carrier to noise ratio (C/N{sub}}0) can drop to levels that exceed the fade margin of the receiver and signal losses are observed. Phase fluctuations are associated with frequency shifts that can exceed the bandwidth of the receiver phase lock loop (PLL) circuit. Both phase and amplitude scintillations are associated with TEC fluctuations. The ionospheric delay caused by the background TEC can be corrected by using dual-frequency receivers, however, short time TEC changes are not so easily corrected and may cause large residuals in the pseudo-range estimation. The IESSG operates a Northern European network of 4 specialised GPS receivers for monitoring of phase/amplitude scintillation, TEC and its rate of change (Dodson et al., 2001). Data collected at those four observation sites during 2002 were analysed to estimate the occurrence of ionospheric scintillation during moderate-to-high solar-flux conditions. The sites are located at Nottingham (52.95°N, 1.18°W, 49.81°N mag. lat), Bergen (60.38°N, 5.26°W, 57.76°N mag. lat), Bronnoysund (65.45°N, 12.47°W, 62.68°N mag. lat) and Hammerfest (70.68°N, 23.71°E, 67.38° mag. lat.). The ionospheric pierce points (ipp) corresponding to GPS satellites being tracked at these sites cover magnetic latitudes from about 40° to 75°. Statistical distributions of phase scintillation magnitudes (based on the 60-sec sigma-phi scintillation index), amplitude scintillation magnitudes (based on the 60-sec S{sub}4 scintillation index) and rate of change of TEC values (TEC change in a 15-sec time bin) are presented in this paper. The statistical distributions were computed for each observation station, each 3-hour time span and for every day of 2002. Results show low occurrence of moderate or strong amplitude scintillation and higher occurrence of strong phase scintillation. Maps of 15-sec TEC changes over Northern Europe and latitudinal profiles of TEC changes have also been generated with results showing that ionospheric TEC disturbances might be observed at magnetic latitudes as low as 50° (350km high ipp) during a major geomagnetic storm. Analysis of strong phase scintillation affecting multiple satellites simultaneously has been conducted. Based on one-year data (2002) we generated a plot that shows the occurrence of multiple satellites being simultaneously affected by different levels of phase scintillation. We also analysed the effects of geomagnetic storms in the occurrence of scintillation in several satellites simultaneously. Analysis of the geomagnetic control of ionospheric scintillation at high latitudes has been investigated. We analysed the correlation between magnetic activity based on the Kp index and occurrence of strong phase scintillation. Results show that scintillation may occur at Hammerfest even during geomagnetic quiet days, however, it seems that only values of K{sub}p ≥ 4 will produce notable scintillation occurrence at Bronnoysund. Finally, we investigated the effects of ionospheric scintillation in different commercial dual-frequency receivers installed in Northern Europe. During a major geomagnetic storm, up to 7 satellites suffered effects from ionospheric scintillation causing loss of phase data (L1 or L2) in the RINEX file. These results are for both codeless and semi-codeless receivers. RINEX files of codeless receivers show many more cases of miss