Jan Kryński, Yevgen M. Zanimonskiy - Monographic Series No 8
Price: 35 PLN
Analysis of variations in the series of GPS solutions and series of gravity observations
The paper presents an original attempt of extension of investigations carried on by research teams on variability of GPS solutions for additional cognitive factors and leads to conclusions of substantial practical use. Quantitative and qualitative analysis of short-period variations of GPS-derived vector components as well as research on their mathematicalmodelling were the main objectives of the project. Modelling of short-period variations of vector components requires an extensive analysis of GPS solutions based on short observing sessions, taking into account correlation with variations of gravity residuals as well as with diurnal and seasonal variability of ionosphere and troposphere. Widely accessible GPS data from EPN stations: Borowa Gora, Borowiec, Jozefoslaw, Lamkowko, Bad Homburg, Dresden, Kiev, Potsdam and Uzhgorad were used in the project. Time series of GPS-derived components of vectors between the permanent GPS stations were analysed. The vectors were computed with the Bernese software as well as with commercial Pinnacle software, using short observing sessions over the intervals varying from two weeks up to half a year. Time series of gravity residuals investigated were obtained from gravity data recorded at Borowa Gora, Jozefoslaw and Lamkowko with computer-controlled LCR-G gravimeters. Meteorological data were also recorded automatically at those stations together with gravity data. Time series of gravity data recorded at tidal stations: Bad Homburg Castle, Black Forest Observatory, Brussels, Membach, Metsahovi, Potsdam, Strasbourg and Wettzell, of the Global Geodynamics Project as well as data from wide-band STS-1 seismometer at Kiev were also analysed to better interpret the variations of gravity residuals especially in terms of their local features. Periodic variations in time series, with periods from a few hours to 24h were, in particular, the subjects of investigations.
In the first step the general characteristics of short-periodic variations in time series of GPS-derived vector components and gravity residuals were investigated. Then the strategy of optimisation of temporal resolution of time series of GPS solutions for detection of periodic biases was developed. It consisted in determination of the optimal length of observing session to ensure a required accuracy for GPS solution as well as in eventual overlapping of observing sessions to provide a sufficient temporal resolution of the time series. Spectral analysis was applied to the time series of GPS solutions of chosen vectors generated with the Bernese software with use of the strategy elaborated. Numerous numerical experiments were conducted to detect short-period biases, separate them taking into consideration their sources and to investigate possibility of their modelling.
Separate chapter in the paper has been dedicated to reliability analysis of estimation of accuracy of GPS solutions with both Bernese and Pinnacle software. The scale coefficients between the parameters of internal and external estimate of accuracy of GPS solutions were obtained on the basis of numerical experiments with numerous time series of GPS solutions.
The use of time series of overlapping GPS solutions makes possible to detect sudden changes in the GPS-derived vector components. Particular attention was paid to the influence of atmosphere on GPS solutions. Time series of Total Electron Content (TEC) in ionosphere, Total Zenith Delay (TZD) in troposphere and meteorological parameters such as temperature, air pressure and humidity as well as of solar activity parameter were generated for time intervals corresponding to time series of GPS-derived solutions investigated. Correlation between short-period variations of vector components and variations of TEC (with special attention to magnetic storms) as well as variations of TZD (with special attention to atmospheric fronts) was investigated. Numerous empirical models that reduce the level of disturbances in GPS solutions were obtained with use of correlation analysis tools.
Variations in time series of GPS solutions were compared with corresponding time series of gravity residual differences generated from data recorded at Borowa Gora, Lamkowko and Jozefoslaw stations. Empirical model for short-period variations of gravity residuals was discussed. The concept of integrated use of time series of GPS solutions and gravity was presented.
The results obtained were used for physical interpretation of short-period variations in GPS-derived vector components and variations of gravity residuals as well as to generate their empirical models.