Coherent Variation in Stress Drop of Small Earthquakes to the Slip Distribution of the 2006 Kiholo Bay, Hawaii, Earthquake
Abstract
We performed multiple-time-window waveform inversion of the MW 6.7 Kiholo Bay earthquake to investigate the rupture speed and slip distribution of the main shock, which occurred on October 15, 2006 beneath the NW part of the Island of Hawaii. We assumed various rupture speeds from 1.0 to 8.0 km/s and evaluated the variance reduction for each assumed rupture speed. Waveforms of an MW 5.2 aftershock were used to calculate empirical Green’s functions. Rupture speeds faster than 3.0 km/s yielded almost the same slip distribution in the corresponding time periods, suggesting that the rupture speed of the main shock was faster than 3.0 km/s (63% of the shear wave velocity). We also found that the rupture propagated unilaterally to the west, in agreement with the location of the main shock epicenter near the eastern edge of the aftershock zone. We also estimated static stress drops of 39 earthquakes (2.5 < ML < 3.5) that occurred in 2006 and 2007 near the source region of the Kiholo Bay earthquake. Static stress drops range from 0.12 and 8.6 MPa and earthquakes around large slip patches of the main shock tend to have larger stress drops. To quantitatively assess this characteristic, we calculated the average value of stress drop in areas with large (area L) and small (area S) main shock displacements. The values were 1.3 and 0.69 MPa for earthquakes in areas L and S, respectively. The p-value of the difference of these two values was 0.13, consistent with our inference that earthquakes around large slip patches of the main shock have larger stress drops. One reasonable possibility is that the patterns of static stress drops of aftershocks reflect coherent variations in the difference between strength and the residual stress level. Unfortunately, there are insufficient data to study the temporal patterns of stress drop before and after the 2006 Kiholo Bay earthquake, because too few earthquakes occurred in the years prior to of the main shock. It is thus still unclear if the stress drop patterns on this fault zone were affected by the main shock. We note as a comparison that Allmann and Shearer [2007] observed largely similar stress drop patterns before and after the 2004 Parkfield earthquake, with only minor temporal changes. Their result may suggest that the spatial pattern of the difference between strength and the dynamic stress level would be stable and would not largely be affected by large earthquakes. Further joint studies of the relationship between main shock slip distributions and the stress drops of small earthquakes may help our understanding of any common processes that control earthquake source characteristics at both large and small scales. In particular, it is important to more fully assess whether the stress drops of small earthquakes can contribute to strong motion prediction by providing information on the locations of large slip regions (asperities) in future large earthquakes.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2010
- Bibcode:
- 2010AGUFM.S43A2026Y
- Keywords:
-
- 7209 SEISMOLOGY / Earthquake dynamics;
- 7212 SEISMOLOGY / Earthquake ground motions and engineering seismology;
- 7215 SEISMOLOGY / Earthquake source observations;
- 7223 SEISMOLOGY / Earthquake interaction;
- forecasting;
- and prediction