The geodetic definition of a stable Eurasian reference frame has long suffered from having a disproportionate number of stations in its European portion and from the fact that the available stations east of the Ural mountains are all in, or close to, the central Asian deformation zone. This has affected, for example, the ITRF2000-Eurasia definition of Altamimi et al. (2002). Only recently this situation has changed with the availability of sufficiently long time-series of IGS stations in central and northern Siberia. These data has effectively been used by Calais et al. (2003) and Steblov et al. (2003). It has become evident that there is a systematic difference between the definition of a Eurasian reference frame from studies that include or exclude stations east of the Ural. The pole of rotation that explains the difference is located in eastern Europe. The rate difference results in up to 4 mm/yr in southeast Asia.

All geodetic studies from the Mediterranean to southeast Asia appear to use different definitions of stable Eurasia. We explore here the effects of adopting different definitions of Eurasia on the geodetic velocity estimates along Eurasia.s southern margin, including Nubia, Arabia, India, and east- and southeast Asia. We show that, compared to Nubia.s motion relative to European stations, its motion relative to Eurasia is 12° more clockwise (CW) in the eastern Mediterranean and 20° more counter-clockwise (CCW) west of Gibraltar. Both solutions are significantly different form geologic estimates, but velocities in a Eurasian reference frame are noticeably closer to the geologic estimates than the velocities in a European definition. Bahrein (Arabia) moves 3° CW and 1 mm/yr faster relative to Eurasia than to a European defined Eurasia. Relative to Eurasia than with respect to Europe. The motion of Bangalore (southern India) could be 3 mm/yr slower and 2° CCW relative to a Eurasian fixed reference when only stations in Europe are used. At Shanghai and Singapore using only European stations would give a velocity 30° CW from a solution that includes Siberian stations. Other examples will be discussed as well.

Finally, it is important to point out that any comparison with geologic plate motion models should be based on a geodetic definition of the Eurasian plate that includes Siberia stations, because the Artic Ridge (which is used in geologic plate motion estimates) extends to 120°E. Also, no-net-rotation (NNR) estimates of Eurasia based on geologic models should be avoided, because its definition is significantly different from geodetic estimates, largely as a result of the inclusion of the Asian deformation zones in the geologic NNR definition.