PM-1

[S06-12] Three-Dimensional Seismic Attenuation Models for Central Nepal

〇Chintan Timsina¹、James Mori²、Masumi Yamada² (1.Graduate School of Science, Kyoto University、2.Disaster Prevention Research Institute, Kyoto University)

We imaged the three-dimensional upper crustal attenuation structure of central Nepal in the source region of the 2015 Gorkha earthquake (Mw 7.8) by analyzing more than 35000 seismic waveforms from local earthquakes. We used P- and S-wave spectra of the Gorkha earthquake sequence recorded by a temporary network of 42 stations deployed for 11 months. We modeled the amplitude spectra for the path-averaged attenuation operator (t*) using a non-linear least square technique by assuming a ω^-2 source model for the frequency band of 1-30 Hz. This procedure depends on finding an optimal fit between observed spectra and theoretically computed spectra based on low-frequency spectral level (Ω_o), corner frequency (f_c), and t*. During the fitting, station site responses are estimated by stacking together all the residual for each station. After removing the site response, the fitting procedure is repeated and the weighting of each t* measurement is also estimated based on how well the spectrum is modeled. We obtained more than 25,000 high-quality t* measurements which are then inverted for frequency-independent Q_p and Q_s structure using SIMUL2000 (Thurber and Eberhart-Phillips, 1999) code. For the ray tracing, we used the 3D velocity models and hypocenter locations from the result of our previous velocity tomography of the region. In general, preliminary results of the attenuation structure agree with the velocity structure, where the high-velocity regions are consistent with high Q-values (low attenuation). In this presentation, we will further discuss the lateral variations on velocity and attenuation structure of the region and their implications for the 2015 Gorkha earthquake.