Novel Application of Receiver Function Analyses Dependent on Splitting Measurement: Crustal Anisotropy along the NAFZ

Abstract

Crustal scale deformation along the fault zone and dipping Moho structures can be constrained by azimuthal seismic anisotropy. Reliable knowledge of the geometry of fault and its vertical extent in the crust and uppermost mantle that often controls observed seismic anisotropy parameters is of great importance for proper seismic hazard assessments in active tectonic settings. The North Anatolian Fault Zone (NAFZ) extending from Karlıova Triple Junction in the east to the Aegean Sea in the west poses actively deforming areas. To elucidate the crustal anisotropy along the NAFZ we will apply a novel receiver function method that simultaneously measures the Moho orientation and average bulk crustal anisotropy. It employs an algorithm in which transverse polarization component minimization (TPCM) applied on the Pms (Moho converted phase) is being integrated into a joint objective method (JOF). The method is advantageous as it restrains the random and coherent noise in the data. Prior to raw data-based anisotropic parameter estimations, we performed synthetic tests mainly considering two hypothetic models devised to be analogous to the NAFZ case. Model-1 assumes S-anisotropy in the crust oriented along N45°E with 4% of strength with flat Moho. The Model 2 involves the same anisotropic properties but with 25° of dipping Moho. Our synthetic tests show that this new approach is able to exactly resolve true model parameters assumed for Model 1 resulting in a 0.987 per cent of model accuracy. The resultant 0.225 s of time delay corresponds to ~4% of anisotropic strength considering a crustal thickness with 30 km and 4.8 km/s average isotropic S-wave velocity for the medium. Our results obtained for Model-2 still tend to converge the true model parameters but show slight discrepancies, in particular, for anisotropic parameters resolved with 0.3 s of time delay and N40°E oriented fast wave azimuth. The test for Model-2 achieves 27.5° as the dipping angle of Moho which is fairly close to its true model parameter. We observe relatively low model accuracy with 0.710 per cent in the case of Model-2. At the further stage of this work, we will utilize digital waveforms of teleseismic earthquakes recorded at the KOERI and AFAD permanent seismic station networks along the NAFZ.