Source Parameters of Intermediate-depth Earthquakes and Application to Early Warning.
Yohei Nishitsuji

We studied various aspects of the source parameters of 106 intermediate-depth earthquakes (MW 3.7 to 7.0) that occurred in the Pacific slab beneath Japan. Using Hi-net data, we made determinations of static stress drop, radiated energy, and radiation efficiency to study the scaling as a function of earthquake size and depth. For determining the source parameters, we also estimated the whole path Q from the earthquakes to the observation stations. Our results show that there is a small increase in the ratio of radiated energy to seismic moment, as a function of seismic moment. This is probably due to an associated increase of static stress drop. Comparisons with the radiation efficiency for shallow crustal earthquakes, show slightly lower values for the intermediate-depth earthquakes. This may indicate that dissipative energy processes may be relatively more important for intermediate-depth earthquakes compared to shallow earthquakes.

Furthermore, we investigated the initiation of intermediate-depth earthquakes to understand which factors are important in determining the size of earthquakes. Our results show that the complexity of the initial portion of rupture for intermediate-depth earthquakes increases as moment increases with constant dynamic stress drop. The estimated duration (around 0.1 to 0.4 seconds) of initial sub-event from the earthquakes which have complex initiation also show constant values as moment increases. These observations suggest that the rupture initiation is not different between large and small intermediate-depth earthquakes. Other factors such as the rupture process or material properties may control the eventual size of the events.

We also investigated ways to apply quick determination of source parameters of intermediate-depth earthquakes for earthquake early warning. The size of earthquakes was estimated using the dominant frequency of the initial portion of the P-wave arrivals. We tested a range of time windows from 1 to 10 seconds for estimating the earthquake size. Our results suggest that using 4 second of first P-wave arrival is adequate for estimating the magnitude for early warning. However, more work needs to be done to improve this methodology for intermediate-depth earthquakes, especially including data for more larger events.