Assessing the accuracy of GNSS VIRTUAL RINEX fill in static positioning.
MetadataShow full item record
In Uganda today, the use of CORS stations is widely being utilized. The use of these stations Other than single base measurements has allowed GNSS users to increase the rover-reference distance beyond the threshold of 20km. Users of SURVNET, and EAGLECORS, can download the static data for processing of GNSS measurements. However, the baseline distances vary from one CORS stations to another depending on the location of the rover receiver. The longer the baseline distance, the more the observation time required to phase out most of ionospheric errors, which mostly affect GNSS measurements as the baseline distance increase beyond threshold. However, this baseline distance can be cut short, by using VRS for static. This phenomenon, allows the creation of the virtual Rinex files (VRFS), in the site by specifying its location within the software, there are new versions of various software that allow this form of processing such as GNSS solutions (3.80.8), Carlson survey GNSS2016 and Trimble (3.5). In this research four known controls(71Ys) and four unknown stations along Lugala Hill road were used which were within 5km.The known controls in this research were first used as unknown stations through creation of the network of VRFs, and also processing using far CORS stations of Entebbe, Luwero and Mubende. For the purpose of knowing the variation of accuracy of VRF with observation time, the data observed was cut to minimum of 30 minutes to 1 hour. Processed and results compared. In both cases VRFs were used as local control bases, and were used to fix both 71Ys controls and stations along Lugala Hill road. The use of near VRS in static succeeded in providing more accurate results than far CORS stations. The average residuals of VRF-processing were below 3cm while the for far CORS, the average residuals exceeded 3cm, from 6 to 7cm, in both Northings and Eastings respectively. Further analysis was conducted using the standard deviations and Root mean square error (RMSE) and far CORS stations produced the highest RMSE of 0.097m. and regarded least accurate. While the use of Virtual Rinex files provided on average of 0.025m Further the observed Rinex file on 71Y82 and 71Y85, were replaced with VRFs in order to process the unknown stations PT1, PT2, PT3 and PT4 respectively. And the processed results were compared. The average of positional difference being below 3cm, and 4cm for 1hour (Real file and virtual file) and 30minute (real and virtual file) respectively.