Absolute Time Series GNSS Point Positioning-Data Cleaning and Noise Characterization

Abstract

Time series data of GNSS point positioning are considerably used for the purpose of geophysical research. The velocity estimates and their uncertainties derive from time series data of GNSS point positioning affected by seasonal signals and the stochastic noise, contained in the series. Data cleaning of GNSS time series is a prerequisite for the noise characterization and analysing. In this article one point positioning of time series was analysed in four different periods during the five year interval. The noise characteristics were estimated for all periods. By applying Lomb-Scargle algorithm the comparable results were also provided. Lomb-Scargle algorithm used to estimate the spectral strength density of unequal sampled data is a typical tool for this kind of analysis. Spectral indices have been estimated before cleaning data and after removing linear, annual and semi-annual signals and outliers. The spectral indices estimated from time series data of GNSS point positioning were located in the area of fractional Gaussian noises, and stationary stochastic process was described for the whole research time period.