Crustal heterogeneity along the Dead Sea Transform - implications to ground motions in Northern Israel

The Dead Sea Transform (DST) dominates the seismicity of Israel, Palestinian Authority, Lebanon and Syria. It is a deep (down to -430 MSL) and narrow (8 to 20 km) morpho-tectonic trough, with a total length of 1100 km. 105 km of left-lateral motion combined with deep sediment filled depressions result in short-wavelength heterogeneity of the upper crust. With displacement rate of < 5 mm/y, the seismicity of the DST is relatively low, resulting in a limited catalog of recorded strong motions.The regional pre-instrumental catalog (> 3000 years), however, contains numerous M > 7. The lack of recorded strong ground motions in Israel results in a critical knowledge gap regarding the distribution of seismic energy throughout the region. This criticality is further amplified given a slip deficit on major segments (equivalent to M > 7) and the proximity of potential sources (< 60 km) to areas of critical infrastructures and urban centers. In this research, we study the influence of the DST crustal heterogeneities on ground motions in northern Israel. The combination of deep sedimentary basins, up to 7 km deep, bounded by active faults is unique and requires in-depth investigation of the impact of this specific setting on the partition of seismic energy in the region. To this end, we have developed an idealized 3-D regional model which includes the DST trough and the north-west branching Carmel Fault Zone (CFZ). The model is based on up-to-date structural data and deep boreholes, includes all the major basins along the DST and the CFZ. The model also includes the Zevulun sedimentary basin, which underlies the Haifa bay area, where dense population and major petrochemical installations are found in proximity. We modeled earthquake scenarios using a kinematic finite fault source and propagated the wave field using the SW4 software. We show that regional energy partition depends on the exact source location along the DST trough, changing from waveguide to resonant box. Concerning the Haifa bay, we show that for similar epicentral distance (55 to 58 km) the amplitude of ground motions is determined by the location of the source along the DST, changing the PGV by a factor 3.This research shows the importance of 3-D numerical models, specifically their ability to address the three components of ground motion convolution: source, path, and site.