Abstract

We compute the secular tectonic motion of the Arabian plate, based on two different sets of Global Navigation Satellite System (GNSS) observations. The first set is formed by continuous GNSS (cGNSS) observations from 10 stations that are being operated since 2000 and have a data time span of at least 3.5 years. The cGNSS is supplemented with a second set of GNSS observations from 11 episodic stations (eGNSS) acquired between 2003 and 2009 (having at least six reoccupations). In order to evaluate the robustness of the solutions, we have estimated three different solutions using different parameterizations concerning the error models as follows: (a) estimating the velocities based only on the cGNSS sites; (b) combining the cGNSS and eGNSS, where the errors for the eGNSS solutions were rescaled to be consistent with those estimated for the cGNSS solutions; and (c) giving the same weight to all cGNSS and eGNSS solutions. Our conclusions show that case (b) is the preferred solution when the uncertainties of the eGNSS solutions are multiplied by a scale factor (4.1x in this particular case), since the inclusion of the eGNSS solutions allows to have a larger number of points with a better spatial distribution. Finally, we compare our angular velocity model with other models, in particular the most recent global models: GEODVEL (geodetic) and MORVEL (geophysical). We find that GEODVEL provides inaccurate predictions for the Arabian tectonic plate, whereas the MORVEL model shows a slightly better agreement with our solution. Our proposedmodel shows a significant improvement to model the present-day kinematics of Arabian plate. Thus, it is suggested to be used instead of the global models. As corollary, we also advocate that global models for present-day tectonic motions should be carefully verified for each modeled plate, so that dedicated regional models can provide better results.