7th International Conference on Earthquake Engineering and Seismology (7ICEES), Antalya, Türkiye, 6 - 10 Kasım 2023, ss.1-7
Abstract. Focal
mechanism solution is important to analyze faulting type, fault geometry, and to evaluate aftershock
patterns after a devastating earthquake occurs. Focal mechanisms were obtained
by S/P amplitude ratios and P wave polarities, are fundamental data to
investigate the geometry of fault zone, slip size, and the crustal stress field.
Methods for solving the focal mechanism conventionally, have categories such as
full waveforms, first motions of P waves, amplitudes of P or S waves, based on
the waveform information used. Due to the difficulty in distinguishing phase
arrivals and first-motion polarities from noise, it is hard to determine the
focal mechanism solutions of small earthquakes. In this study, to overcome this
difficulty, we applied artificial
intelligence to solve the focal mechanisms of small earthquakes in Izmir and its surroundings, where there is
high microearthquake activity. The data set required for artificial
intelligence applications was generated by using the records of earthquake
stations located in western Turkey,
especially in Izmir and its surroundings. Focal mechanism solution for Buca
earthquake was determined by using Artificial neural
networks are a branch of artificial intelligence.
We compare the results obtained with artificial intelligence and traditional
methods, and it is seen that artificial intelligence can be applicable for the
focal mechanism solutions.