Occurrence of intraplate earthquakes have been related either to reactivation of preexisting zones of weakness in a uniformly oriented regional stress field or to a local stress changes. More recently, buildup of pore pressure anomalies has been proposed to explain both the low recurrence rate and some misorientation of rupture planes in the prevailing regional stress field. In this study, we analyze local stress field orientation in the area of the three M(s) = 7.0 Gazli (western Uzbekistan) earthquakes by using different sets of earthquakes focal mechanisms. We show that stress tensor inversions from both seven 4.2 less than or equal to M(s) less than or equal to 7.0, forty-five 0.1 less than or equal to ML less than or equal to 3.2 and twenty-five 1.0 less than or equal to ML less than or equal to 3.2 earthquake focal mechanisms give constant maximum horizontal stress direction in agreement with the one deduced from the India-Asia collision. The best model for the local stress tensor exhibits a significant (25 degrees) discrepancy from vertical for one of the principal stresses. We note, moreover, that the common assumption that one principal stress axis is oriented vertically is not strictly supported by stress states derived from earthquake focal mechanisms available in the world stress database. By evaluating each hypothesis of stress inversion methods from earthquake focal mechanisms, it is shown that, especially in intraplate settings, the first motion focal solution should be the most appropriate one for a formal inversion and that the magnitude of the considered event may not be the only criterion to assess the quality of a stress inversion. This is particularly timely because more and more focal mechanisms are moment tensor solutions instead of first motion solutions. Concerning the Gazli sequence, the slip conditions (pore pressure and frictional coefficient) are evaluated for each of the three mainshocks in terms of an independently determined local stress field. Because variations in orientations between local and regional stress field are not detected, and because the three mainshock fault planes seem to be relatively well oriented for a frictional activation in the local stress field, it is concluded that this seismic sequence is localized on preexisting weakness zones and is not due to local stress orientation anomalies.
