INVESTIGATION OF THE ACCURACY OF POST-PROCESSED STATIC GNSS DATA OF KNOWN POINTS USING CONSTELLATION DISCRIMINANT AT UNIVERSITY OF BENIN, UGBOWO CAMPUS, BENIN CITY

This study investigates the influence of satellite constellation configurations on the positional accuracy of post-processed static GNSS data at the University of Benin, Ugbowo Campus. Static observations were collected at five known control points using a Tersus David30 receiver. Data was processed in Tersus Geomatics Office across seven constellation setups: GPS-only, GLONASS-only, BEIDOU-only, and their combinations. A detailed epoch-based analysis was also conducted at one control point using RTKLIB. Accuracy was assessed using coordinate residuals (∆E, ∆N, ∆H), RMSE, standard deviation, CEP, and 2DRMS, supplemented by classical
and robust statistics and time-series analysis. Results demonstrated that GPS-based solutions consistently delivered superior performance. The GPS+BEIDOU combination achieved the best accuracy (2DRMS = 0.160 m, CEP = 0.067 m), closely followed by GPS-only. In contrast, BEIDOU-only yielded the poorest results (2DRMS = 0.587 m), while GLONASS-only was notably weak and unstable. RTKLIB processing confirmed that multi-constellation setups, particularly GPS+GLONASS+BEIDOU, produced highly precise solutions with sub-centimeter standard deviations. Conversely, the GLONASS-only solution exhibited severe instability, with significant errors and outliers. Time-series analysis revealed that stable constellations maintained narrow error bands, while error spikes in other configurations corresponded directly to drops in satellite visibility