Selection and Modification of Ground Motion Prediction Equations in Different Tectonic Regions of Iran considering Declustered and Non-declustered Earthquake Catalogs.

Shallow crustal earthquakes (SCE) occur in different seismotectonic regions of the Iranian plateau. In this study, 10 ground motion prediction equations (GMPEs) developed based on local, regional, and global data from SCE are considered as candidates to test their explanatory power against Iranian strong motion data. To show the applicability of candidate GMPEs in the plateau as a united region and its different seismotectonic sub-regions, a dataset including 3045 records from 678 seismic events (both mainshocks and aftershocks) is used for statistical analyses. Moreover, using 1662 records from 245 mainshocks, we tried to find more appropriate GMPEs against mainshock ground motions through statistical tests. This work also concentrates on differences in ground motions from mainshock and aftershock earthquakes using the Iranian database. The statistical analyses including LH (Likelihood), LLH (Log-Likelihood) and EDR (Euclidean distance-based ranking) are also performed on GMPEs modified by a local anelastic term. According to the LH and LLH analyses results, local and regional models have better performances for PGA and spectral accelerations (SA) at different periods ranging from 0.05 to 3 sec. However, according to the EDR test, it can be said that some global GMPEs also show acceptable performance for predicting the Iranian ground motions. Also, the adjustment factors added to the GMPEs can considerably improve their explanatory performance. [ABSTRACT FROM AUTHOR]