[S02-P] Poster

[S02-P-06] Searching significant displacement zones of a M5.5 earthquake fault by forward and inversion analyses of strainmeter data at depth at a very close distance

^○Tatsunari Yasutomi¹, Hiroshi Ogasawara², Akimasa Ishida², Hiroyuki Ogasawara², Durrheim Raymond³, Alex Milev⁴, Makoto Okubo⁵, Teruhiro Yamaguchi⁶, James Mori¹ (1.Kyoto University, Kyoto, Japan, 2.Ritsumeikan University, 3.Witwatarsrand University, South Africa, 4.CSIR, South Africa, 5.Kochi University, 6.Hokkaido University)

The largest event recorded in a South African gold mining region, a M5.5 earthquake, took place near Orkney, South Africa on 5 August 2014. It is an unusually well-recorded events: e the main- and after-shocks were recorded by 46 geophones and 3 Ishii borehole strain meters at 2 - 3 km depths with epicentral distances, delta < several km, and 17 surface strong motion meters with delta < 20 km. The upper edge of the planar aftershock activity dipping almost vertically was only some hundred meters below the sites where the strainmeters were installed. As the M5.5 seismic rupture is located within a distance that is drillable from the deep gold mine workings, ICDP approved a project to drill into the seismogenic zones. Moyer et al. (2016 SCEC) inverted the surface strong motion data, suggesting that there was significant fault slip, even at the mining horizon. However, no seismic rupture was mapped in the mine workings. So, the three strainmeters can constrain the configuration of the seismic rupture. As the population of the aftershocks varies significantly in space, we will discuss the relationship the fault slip and the aftershocks.
These strainmeters were only about 150 m o apart. However, their strain changes had different polarities while the other M4 strain changes with a similar hypocentral distance was the same. So, this information can place critical constraints on the location and configuration of the M5.5 fault.
First, we conducted a forward analysis by assuming a point source with the mechanism same as macroscopic one of the M5.5 faultingat a distance of a few km. However, no difference in the polarity of strain change was seen, suggesting that the effect of a finite size of the source with an edge much nearer than the point source had to be taken into account. We are attempting to invert the slip distribution on a source with a finite size together with surface strong motion data. We will report on the results at the meeting.