Thursday 25 August 2005

G1
1300-1500 hours

386
A comparison of BIFROST products
Bergstrand, Sten¹, Lidberg, Martin¹, Scherneck, Hans-Georg¹, Johansson, Jan M¹
1 Onsala Space Observatory, Chalmers, Onsala, Sweden
Author email: sten@oso.chalmers.se
This study compares the end products from different GNSS analysis softwares and examines the effect of a common-mode removal from single-site solutions in the BIFROST network. The BIFROST (Baseline Inferences for Fennoscandian Rebound Observations, Sea level and Tectonics) GPS network of more than 30 stations has been used to observe and constrain parameters connected to glacial isostatic adjustment since 1993. Subsequent to 1996 no major changes in the network have been adapted, and we are now reaching a state where a comparison between different analysis software packages and strategies is viable. We compare recent and earlier processing results from eight years (1996-2004) of continuous GPS data processed with both GIPSY-OASIS II and GAMIT-GLOBK, and analyze the stochastic properties of the stations' time series residuals. The results are used to gauge uncertainties of intraplate movement, which in our case typically are on the order of less than 1 mm/yr. We compare the results to previously estimated observation uncertainties used for inferences of geophysical parameters. The new solutions appear less noisy, and thus provide means to give more accurate estimates of the station motion. The results are also useful to estimate regionally coherent systematic noise. Spectral analysis of the residuals and the regional common mode generally show fractal noise characteristics (power spectra are proportional to the time series frequency components to a power between -0.5 and -0.9). At low frequencies, however, the slope of station residuals is seen to flatten out, which indicates more stable long-term conditions in the regional network than in the common mode.

Return to Oral Presentations