Modelling the future Israel EEWS performance using synthetic catalogue

In 2012 June, Israel government decided about building an Earthquake Early Warning System (EEWS) in the country. The network configuration suggested was to be a staggered line of ∼100 stations along the regional main faults: Dead Sea fault and Carmel fault and additional ∼22 stations spread over the country. The EEWS alarm system should have to utilize two approaches: the P-wave-based algorithm combined with the S-threshold method. The former utilizes first wave arrivals to several closest stations for prompt location and the three initial seconds of the waveform data for magnitude estimation. The latter issues alarm when the surface shaking (velocity or acceleration) exceeds the relatively high threshold corresponding to a magnitude 5 earthquake at a short distance of about 10 km at least for the two neighbouring stations. For each of the approaches and for a reasonable combination of them, we simulate the EEWS performance based on a synthetic catalogue. The input seismicity parameters for the processing are extracted from the real instrumental catalogue. Using a general ground-motion prediction equations for peak ground accelaration, τc and Pd we then evaluate how false and missed alarm rates depend on the corresponding thresholds. Practically, in turn, these dependencies approve choosing initial thresholds for the EEWS providing suitable false and missed alarms rates.