Optimal modelling of position errors with respect to observation duration for online GNSS processing services

In this thesis, the positioning capabilities offered by GNSS (Global Navigation Satellite System) technology were evaluated through online data processing services, and the relationship between measurement time and positioning accuracy was examined. Commonly used internet-based services such as AUSPOS and CSRS-PPP were analysed in comparison with the scientific software GAMIT-GLOBK. While the AUSPOS service does not provide solutions for measurements shorter than 1 hour, the CSRS-PPP service provides results for all measurements longer than 5 minutes. Accuracy, repeatability, and differences between software were investigated in the horizontal and vertical components using a total of 250 static GNSS data sets collected at different intervals and repeatedly. The study concluded that position accuracy is directly related to observation duration, with longer observation periods yielding more meaningful results, particularly in the vertical component. While centimeter-level accuracy can be achieved with 1–3 hour observations in the horizontal component, 24-hour observations are required for milimeter-level accuracy in the vertical component. Additionally, coordinate differences between software were analysed, and systematic deviations were identified, particularly in the vertical component. The findings contribute to the development of error models that will assist users in making informed decisions regarding measurement duration and service selection when planning GNSS observations.