14:45 〜 15:00
[SCG70-05] Characteristics of Near Fault Strong Ground Motion in the Kathmandu Valley during the 2015 Gorkha Nepal earthquake
キーワード：Fling-Step, permanent tectonic offset, Site Amplification
On 25 April 2015, a large Mw 7.8 earthquake occurred along the Main Himalayan Thrust fault in central Nepal. The epicenter was near the Gorkha region, 80 km north-west of the Kathmandu Valley, and the rupture propagated eastward from the epicentral region passing through the Kathmandu Valley. We have installed a strong motion array observation (four sites; one rock site and three sedimentary sites) in the Valley, on 20 September 2011, to understand the site effects of the Valley. We discuss the characteristics of near fault strong ground motion in the Kathmandu Valley during the 2015 Gorkha earthquake based on the strong motion records captured by this array.
The horizontal velocities waveforms on sedimentary sites are strongly affected by site amplification due to soft soil deposit and valley response (Takai et al. 2016; Galetzka et al. 2015). The velocity waveforms for the N207E (fault normal direction) and UD components observed at the rock site KTP show the distinguishing velocity pulse ground motions. They show a single-sided velocity pulse with a width of about 6 s, while the N117E (fault parallel direction) component show a double-sided pulse with a period of about 10 s. This N117E pulse is considered to be effect of along-strike directivity (Mavroeidis and Papageorgiou 2003), and the pulse shape was explained by the joint inversion result for rupture process (Kobayashi et al. 2016). The ground velocities at KKN4 obtained from the high-rate (5 Hz sampling) GPS record (Galetzka et al. 2015) have the similar waveforms as observed at KTP, while the amplitudes of the KKN4 velocity pulses are about 1.4 times larger than those of the KTP velocity pulses; KKN4 is a rock site located northwest of the Kathmandu Valley.
The Kathmandu Valley is located at a very close distance (~10 km) to the rupture area and the estimated large slip areas exist near the Valley. Furthermore, the displacement waveforms derived from the velocity pulses for the N207E and UD components at KTP show a monotonic step. These facts demonstrate the observed velocity pulses are effect of a permanent tectonic offset (Mavroeidis and Papageorgiou 2003). If the records are affected by the permanent tectonic offset, the velocity waveforms are similar to the slip-rate functions (Hisada and Bielak 2003). Galetzka et al. (2015) estimated the regularized Yoffe slip-rate time function from the vertical velocity waveform at KKN4 by forward modeling. They also showed that the estimated slip-rate time function well explained the vertical velocity waveforms at two stations in the Kathmandu Valley. We confirmed that the estimated slip-rate time function well explained the vertical velocity waveforms at our four stations. It is interesting to extract the slip-rate time function from the observed records without waveform modeling. We made a trial of extraction of the slip-rate time function based on the low-pass filtered acceleration waveform for vertical component at KTP; the cut-off frequency of the filter is 0.3Hz. The velocity and displacement waveforms obtained by single and double time-integration of the low-pass filtered accelerations show the velocity pulse and the monotonic step, respectively. We also confirmed the Fourier spectral shape at the low-frequency range (0.02-0.3Hz) of our low-pass filtered acceleration waveform is similar to that of the differentiated Yoffe slip-rate time function estimated by Galetzka et al. (2015).
Kamai et al. (2014) developed an empirical parametric model for the fling-step components based on an extensive set of finite-fault simulations. We compare the width of the velocity pulse (about 6 s) observed at KTP with their regression model of the period of the fling-step pulse for the reverse fault. The width of velocity pulse at KTP, is nearly the same as the median value of the regression model by Kamai et al. (2014). This means that the Gorkha earthquake with Mw 7.8 is normal one with respect to the fling-step motion.
The horizontal velocities waveforms on sedimentary sites are strongly affected by site amplification due to soft soil deposit and valley response (Takai et al. 2016; Galetzka et al. 2015). The velocity waveforms for the N207E (fault normal direction) and UD components observed at the rock site KTP show the distinguishing velocity pulse ground motions. They show a single-sided velocity pulse with a width of about 6 s, while the N117E (fault parallel direction) component show a double-sided pulse with a period of about 10 s. This N117E pulse is considered to be effect of along-strike directivity (Mavroeidis and Papageorgiou 2003), and the pulse shape was explained by the joint inversion result for rupture process (Kobayashi et al. 2016). The ground velocities at KKN4 obtained from the high-rate (5 Hz sampling) GPS record (Galetzka et al. 2015) have the similar waveforms as observed at KTP, while the amplitudes of the KKN4 velocity pulses are about 1.4 times larger than those of the KTP velocity pulses; KKN4 is a rock site located northwest of the Kathmandu Valley.
The Kathmandu Valley is located at a very close distance (~10 km) to the rupture area and the estimated large slip areas exist near the Valley. Furthermore, the displacement waveforms derived from the velocity pulses for the N207E and UD components at KTP show a monotonic step. These facts demonstrate the observed velocity pulses are effect of a permanent tectonic offset (Mavroeidis and Papageorgiou 2003). If the records are affected by the permanent tectonic offset, the velocity waveforms are similar to the slip-rate functions (Hisada and Bielak 2003). Galetzka et al. (2015) estimated the regularized Yoffe slip-rate time function from the vertical velocity waveform at KKN4 by forward modeling. They also showed that the estimated slip-rate time function well explained the vertical velocity waveforms at two stations in the Kathmandu Valley. We confirmed that the estimated slip-rate time function well explained the vertical velocity waveforms at our four stations. It is interesting to extract the slip-rate time function from the observed records without waveform modeling. We made a trial of extraction of the slip-rate time function based on the low-pass filtered acceleration waveform for vertical component at KTP; the cut-off frequency of the filter is 0.3Hz. The velocity and displacement waveforms obtained by single and double time-integration of the low-pass filtered accelerations show the velocity pulse and the monotonic step, respectively. We also confirmed the Fourier spectral shape at the low-frequency range (0.02-0.3Hz) of our low-pass filtered acceleration waveform is similar to that of the differentiated Yoffe slip-rate time function estimated by Galetzka et al. (2015).
Kamai et al. (2014) developed an empirical parametric model for the fling-step components based on an extensive set of finite-fault simulations. We compare the width of the velocity pulse (about 6 s) observed at KTP with their regression model of the period of the fling-step pulse for the reverse fault. The width of velocity pulse at KTP, is nearly the same as the median value of the regression model by Kamai et al. (2014). This means that the Gorkha earthquake with Mw 7.8 is normal one with respect to the fling-step motion.