Discrimination of the natural and artificial quakes in the Eastern Marmara Region, Turkey

SERTÇELİK F. , Yavuz E. , Birdem M., Merter G.

ACTA GEODAETICA ET GEOPHYSICA, cilt.55, sa.4, ss.645-665, 2020 (SCI İndekslerine Giren Dergi) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 55 Konu: 4
  • Basım Tarihi: 2020
  • Doi Numarası: 10.1007/s40328-020-00315-3
  • Sayfa Sayıları: ss.645-665


The aim of this study is to determine the source of seismic quakes is whether natural or artificial in the Eastern Marmara Region which is also included two branches of western part of the North Anatolian Fault Zone. It is known as one of the most seismically hazardous region in Turkey and 1999 Izmit and Duzce Earthquakes were occurred inside this area. In addition, it has geothermal springs, lots of mining areas, tunnel, bridge and road construction points. Under these conditions, it is seismologically important to make a study about the discrimination of the natural and artificial events. 611 seismic events with Ml <= 2.5 which has occurred between 2017 and 2019, recorded by 11 seismic stations that are being operated by Republic of Turkey Prime Ministry Disaster and Emergency Management Authority Presidential of Earthquake Department were analyzed. Amplitude ratio, complexity and continuous wavelet transform (CWT) methods were performed; linear and quadratic discriminant functions (LDF, QDF) were used for statistical processes to all waveforms and their results. All waveforms were identified by significant manual investigation that P wave first motion polarity, no or low amplitude S wave recording, the vision of coda wave decay and Rg phase detection and it called as a first determination. The results were compared to first determination and the final decisions were gained with the combination of their all. Besides with the all methods, CWT analyze could give more reliable results than the others and amplitude ratio method can serve better results than complexity. Otherwise, LDF could classify the events better than QDF. Through the final decision, 395 earthquake and 182 blast recordings were clearly discriminated with the accuracy of 94.44% and the source types of 34 seismic events could not be identified. The obtained discriminant functions for each station are useful to determine the source types of the events and they could clarify the contaminated earthquake catalogues. Increasing the reliability of the final results would be related with using these methods together.