Towards a regional, automated full moment tensor inversion for medium to large magnitude events in the Iranian plateau

Abstract

Considering the high seismicity rate of the Iranian plateau, the development and implementation of an automated routine for near-real-time determination of earthquake point source parameters would be extremely beneficial, e.g., for dedicated seismological studies, constraining the regional stress field, and shakemap assessment. We discuss the implementation of a fast, automated regional moment tensor inversion routine, based on SeisComP3 and KIWI tool packages. To assess the routine efficiency, we used a dataset including 90 earthquakes spatially distributed within all seismotectonic provinces of Iran, recorded at local to regional distances, with magnitudes in the range MN 4.5–7.3. Resolved moment tensor solutions and source parameters are in most cases in good agreement with reference ones from global and regional catalogs. However, we identify consistent differences in magnitudes and depths, with respect to global catalogs, which can be attributed to a higher resolution of our approach, using local and regional data and considering regional velocity models. The depth estimations are shallower than in reference global catalogs but compatible with the tectonic settings of the Iranian region. The analysis of the selected dataset is accompanied by a test run of automated processing for 8 months of online operation, from July 2019 to February 2020. The processing time in online routine, with an average of ~ 23 min, mostly depends on the magnitude, number of stations, azimuthal coverage, data quality, and complexity of the crustal model, so that computation times may vary significantly for earthquakes located in different Iranian provinces.