A Unified approach in GPS accuracy determination studies

Abstract

By the time GPS technology started to be used in Geodesy, it is much easier to reach the desired precision of point positioning. It is significantly a useful technique, thus one can easily predict the accuracy of GPS before a field survey and know about the quality of the observations that have been made on a reference point. Parallel to the improvement of the GPS technology, predicting the accuracy over short and long baselines has really been an important discussion. There have been several studies dealing with precise point positioning and the topic was to determine how the accuracy depends on the baseline length and the duration of the observing session (Eckl et al., 2001, Soler et al., 2005, Do an, 2007, Engin and Sanli, 2009). In the previous studies, the accuracies for the baselines were taken into account separately, and models have been created for the baselines between 30-300 km and 300- 3000 km. For the baselines smaller than 300 km, the accuracy was found to be a function of only the observing session duration (Eckl. et al., 2001) but for the baselines between 300-3000 km the results show that it does not only depend on the observing session it also depends on the inter-station distance (Engin and Sanli, 2009). In this study, the aim was to make the discussion topic certain and to combine a model for baselines ranging from 3 km to 3000 km. To define a unified model, GPS accuracy was tested in IGS network and the results are compared with recent studies by using GIPSY software. 13 baselines and the data of 10 days have been used in this research. Baseline lengths were between 3 km and 2739 km. The data of each day have been divided into sub sessions (6-8-12 and 24 hours) and then evaluated separately. Thus, the relation among GPS point positioning, base length and duration of observation has been examined. The results show that, the point positioning accuracy in IGS network over 3-3000 km depends both on the baseline length and the observing session duration. It is partially possible to define a unified model for baselines between 3 and 3000 km. To define a unified model for this range, could only be possible by testing out the significancy of various sub sets of Least Squares coefficients.