Quality Structure of Time Series of GPS Solutions

摘要

Growing interest is paid in the community of GPS users to variations in positions or vector components obtained from processing precise GPS observations. The observed phenomenon needs to be investigated to improve modelling of GPS observations for precise positioning and to develop spatially extended real-time GPS technology. Variations in GPS solutions have a complex structure. they can be represented in terms of biases as well as components generated by both random and chaotic processes. Short-term variations, i.e. diurnal and sub-diurnal are of particular importance for quick and efficient precise positioning. Regular terms with distinguished periodicity dominate in those variations. Variations in GPS solutions are caused by numerous factors. Thus their quantitative estimation that is necessary for further more adequate modelling of GPS observations should be preceded with qualitative analysis. Data from permanent GPS station arrays is, in particular suitable for such investigations. A method for increasing time resolution of GPS solutions with simultaneous reduction of random errors has been developed. Numerous numerical experiments have been conducted to separate biases caused by different factors.