Velocity model calculation and seismicity study of last decade on Tehran and high Alborz elevations

An earthquake of local magnitude 5.2 which occurred in Malard recently, was accompanied by four earthquakes with local magnitudes largerthan 4.0 These events can be considered as a manifestation of the state of Tehran, a mega-city with more than 13 million inhabitants, in last decade. For this reason, we tried to do a comprehensive seismic study on Tehran and surrounded area using all available data. In order to do this, we used three independent datasets belonged to Iranian Seismological Center (IRSC), International Institute of Earthquake Engineering and Seismology (IIEES) and Tehran Disaster and Mitigation Organization (TDMMO). We merged these three datasets to get a uniform one, using 3D cells gridding technique and then explosions or outliers were removed. The final dataset included 2000 events. Using 310 selected events based on horizontal and depth errors, azimuthal gap, RMS and minimum number of recording stations, we calculated 1D velocity model by utilizing Particle Swarm Optimization (PSO) method. The calculated model consists of three layers. The thickness and P velocity of layers are 4.0 km and 5.4 km/s, 6.0 km and 5.8 km/s and 5.0 km and 6.05 km/s for the first, second and third layer, respectively. The final layer is a half-space with a P velocity of 6.5 km/s. The computed Vp/Vs is 1.71. Then all events were relocated using our new velocity model utilizing fully non-linear probabilistic method to get as much as possible accurate locations. The results show that 40% of all relocated events have uncertainties less than 2.5 km and 5.0 km in horizontal and vertical direction, respectively. The final calculated mean RMS is ~ 0.24 s. In order to define the geometry of the active faults, three different subsets of events were selected based on their location uncertainties, azimuthal gap, RMS and minimum number of recording stations. This helped us to use well-located events for better defining of the fault traces on map view and in depth cross-sections. We plotted four cross-sections perpendicular to the strike of the main faults in the region. The focal mechanism solutions for 38 selected events were also computed based on P-wave first polarity method. The final results show that the eastern part of the study region is more active than the western part, at least in the last decade, and surprisingly, the Malard earthquake occurred in a region without any major activity from three months before the main shock. Stress field study also reveals that the maximum stress axis is N36E and the main seismotectonic regime is left lateral structures. These results are very consistent with the main trends of the North Tehran fault