Continuous evolution of Global Navigation Satellite Systems (GNSS) provides the opportunity of using observations from multi-GNSS constellations for geodetic applications. It is a well-known fact that the use of multi-GNSS observations improves the robustness and reliability of the position estimates while providing enhanced spatial coverage due to increased number of available satellites. The combined use of the ground-based observations from GPS and GLONASS, for example, significantly adds to the coverage in both azimuth and elevation of the ground-based stations. The aim of this study is, firstly, to assess the geographical variations in the improvement (of position estimates) offered by the use of multi-GNSS observations, and secondly, to assess implications of multi-GNSS position estimates for improvements and robustness of the terrestrial reference frame. A GPS-only double-differenced positioning solution for a global network of stations will be used as a reference in this study. Furthermore, two precise point positioning (PPP) solutions i.e. one with GPS-only and one with GPS+GLONASS observations, will be computed and compared to the reference solution. To study the geographical variation, the differences between the simulated and reference North/East/Up components will be analyzed as a function of latitude. Eventually, the various positioning solutions will be analyzed in terms of their application in terrestrial reference frame definition.