Comparison of Loading Models, GLONASS and GPS Systems in Series Station Positions

V. L. Gorshkov, N. V. Scherbakova

Transactions of IAA RAS, issue 37, 68–71 (2016)

Keywords: GLONASS/GPS position series, GNSS observational errors, atmospheric and hydrological loading models.

Abstract

The daily position series of selected GNSS stations around the Golf of Finland were calculated by PPP strategy of Gipsy software separately for GPS and GLONASS systems. The free solution and IGS and ESA orbits and clock corrections were used for comparison of GNSS systems. The data only after 2009 year were used for that as the GLONASS satellite system was mainly completed in this year. But all station data were used for the comparison of the loading models from IMLS (http://massloading.net). The systematic errors of the investigated site position (seasonal and more low-frequency) and noise estimations (mean error relative to linear regression σy and weighted Allan variation WADEV) were compared for various loading models and GNSS systems. There are no differences in noise estimations between atmospheric loading model influences whereas its seasonal variations (with 1–3 mm amplitudes) are different sometimes and partly compensate the seasonal variation of the initial position series. Both noise estimations are significantly improved for vertical components of the site positions by using any atmospheric loading models. The spectra of land water storage mass loading (LWS) models are mainly seasonal and very different for various models especially before 2011 year. The most homogeneous of them is MERRA. Neither seasonal errors nor noise estimations of initial position series improve by using any LWS models. One sigma errors of the daily site position are comparable for horizontal components calculated in both GNSS systems whereas vertical components are significantly better in only GPS solution. The same conclusions are for the WADEV estimations. The small seasonal variations are remained in the differences between site positions calculated in both GNSS systems.

Citation