DOI: 10.5593/SGEM2014/B22/S9.067


J. Bogusz, M. Figurski, A. Klos, A. Araszkiewicz
Wednesday 1 October 2014 by Libadmin2014

References: 14th International Multidisciplinary Scientific GeoConference SGEM 2014, www.sgem.org, SGEM2014 Conference Proceedings, ISBN 978-619-7105-11-7 / ISSN 1314-2704, June 19-25, 2014, Book 2, Vol. 2, 537-544 pp

The stochastic part (noises) of the GPS-derived time series has a direct impact on the uncertainties of the linear parameters estimated from this data. The noises are commonly analysed with the Maximum Likelihood Estimation (MLE) that is stated as being the most precise method for this purpose. Furthermore it is a very powerful method for different practical research e.g. investigation of permanent GNSS station’s stability. In general, the noises are recognized as the power-law processes characterized by the spectral index κ and the amplitude A. There are three integer values of spectral indices, that stand for the special cases of noises: white noise, flicker noise and randomwalk. The most of the papers showed, that noises in the GPS time series are the closest to flicker one, which is the effect of mismodelled satellite orbits, Earth Orientation Parameters or large scale atmospheric influences. Besides, the amplitudes of noise reveal the latitude dependence for vast networks. In this research, the daily changes of topocentric coordinates (North, East and Up components) in the ITRF2005 from the set of EPN (EUREF Permanent Network) stations were used. The time series were obtained by the Bernese 5.0 processing (“repro1” project) performed in the Centre of Applied Geomatics that cooperates at the Military University of Technology as one of the 16 EPN Local Analysis Centres (MUT LAC). The time series were pre-analysed with the median absolute deviation (MAD) criterion and sequential t-test algorithm (STARS) to remove outliers and offsets. The stochastic part of GPS time series was analysed with the MLE method with white plus flicker noise model assumed a-priori. As the result, the amplitudes of the aforementioned noises for all of the EPN stations were estimated. The EPN network is located between the meridians of 27° N and 80° N, what gives the 53 degrees of latitude. The paper presents the comparison of noise amplitudes in relation to the geographical location of permanent GPS stations.

Keywords: GPS, EPN, noise analysis, latitudinal dependence