The Relocation of April 5th. 2017 Sefid-Sang Earthquake aftershocks sequence using Waveform cross-correlation and double-difference techniques

On April 5, 2016, the Sefid-Sang earthquake with a magnitude of Mw~6 occurred about 30 km from Sefid-Sang and 80 km from Mashhad in Khorasan-Razavi province, Iran. Looking at the distribution of aftershocks reported by the Iranian Seismological Center (IRSC), due to the error in the locating of the earthquakes, it is not possible to attribute these events to the activity of any pre-mapped faults. In this study, we attempt to determine the relative location of earthquakes and improve the seismic phase readings on the data recorded in the seismic waveforms recorded by seismic stations of the IRSC to find the causative fault. To relocate the aftershocks of the Sefid-Sang earthquake of April 2016, 2136 data that were recorded in the period from April 2016 to May 2017 at IRSC, were obtained from this center. First, a spatial distribution was drawn for these data, and as said before it was very difficult to determine the causative fault. The data were relocated using the double difference method and also apply the cross-correlation on the waveforms to improve the reading of the seismic phases. In this research, the aftershocks relocated in four different cases. Once the entire data without improving seismic phases relocate with HypoDD together, the next step, first using cross-correlation on all waveforms, we improve the phase readings, and then HypoDD is used to relocate all aftershocks. Then we try to use this double difference method on the data with a better condition in the first reported location. We selected data with an azimuth gap less than 180 degrees, RMS less than 0.5 seconds, and magnitude less than 2.5. Again on selected data, HypoDD was run in two different ways, once only using the phases which read by IRSC experts, and the second, addition, of the cross-correlation output of the waveforms was used to improve the readings. By examining the depth sections on two profiles, first in the northwest-southeast direction and the second, perpendicular to the first profile, it was found that the dip of the fault plane is about 45 degrees towards the northeast, and accordingly, the fault direction is northwest-southeast. Using InSAR image processing, Ghayournajarkar and Fukushima, 2020, stated that aftershocks are more compatible with the fault model whose dip is northeast, and the dip of the fault model has a dip of 47.4 degrees. As can be seen, most of the studies conducted, such as the result of this study, consider a fault with a northwest-southeast strike and northeast dip as the fault that caused the 1396 SefidSang earthquake. According to the existing faults in the region, this fracture is the most consistent with the continuation of the Kashafroud fault. Considering the importance of the Kashafroud fault in determining the seismic hazard of Mashhad city, this shows the necessity of studies to better identify this fault zone and its parallel branches. Compared to the study of Einakchie et al. 2018, this study showed that to obtain favorable results in determining the relative location of earthquakes, efforts should be made to improve the seismic phase readings by any method such as waveforms cross-correlation. Also, despite not having the data of the temporary local dense network used by Khosravi et al. 2018, by choosing stricter conditions on data selection and also improving the reading of seismic phases, and determining the relative location to reduce the error caused by the velocity model, the results could improve a lot, of course, in this study the main advantage is the existence of the JRKH permanent station less than 10 km from the epicenter of the mainshock. The importance of the temporary seismographic network around the epicenter of the earthquake increases the number of records and increases the accuracy of earthquakes, but in cases in which we do not have access to such networks data, existing of the close station to epicenters and using double-difference method could help to get more accurate results.