A PRELIMINARY REPORT OF INVESTIGATIONS
ON SOUTHERN HYOGO PREFECTURE EARTHQUAKE

Research Center for Earthquake Prediction
Disaster Prevention Research hstitute
Kyoto UniversityiRCEP-DPRIj

When tle big earthquake measurig 7.2 in Richter magnimde scale occurred in Kobe,Csouthern Hyogo prefecture on 17 th of January 1995,12 seismometes of Abuyama Observatory, an observatory to our research center(RCEP-DPRI), were in place and oprating near and around area of the earthquake's epicenter. Immedietely after the said event, these seismometers recorded numerous afershocks whose hypocenters were automatically determind by a computer in Abuyama Observatory. The main shock was located around Akashi straight, several kilometers north of Awajshima island, westem Japan.

FigDI Focalmechanisms ofthe 1995 South Hyogo Earmquake and its foreshocksD

We found four foreshocks associated with this earthquake with epicenter located at Akashi straight almost in the same zone of the main shock and occurred approximately 12 hours and 6 hours after the main shock. Focal mechanisms of the main shock and those of the foreshocks are shown in Fig.1. These focal mechanisms indicate typical strike-slip motion with right-lateral displacement. This motion is due to the east-west tectonic compression which westem Japan expriences at present.

From January 17 to February 14, 1995, the total number of recorded aftershocks is more than 6000. Hourly count of aftershocks, however, gradually decrased and fits the aftershock decay curve defined by the Omori formula(Fig.2). During the same observation period, the largest aftershock(M4.9) was recorded about 2 hours after the main shock. The epicentral map of aftershocks from January 17 to February 14 is shown in Fig.3. The space-time distribution of afershocks along the AB line in Fig.3 is shown in Fig.4.

Results of our preliminary analysis of the aftershock data, the January 17 earthquake was due to the movement of the fault passing at Akashi straight. This site of the hypocenter of the main shocks presently characterized by a relativelyl low aftershock actrivity(Fig.4). Consequently and probably within several seconds after the motion in the hypocenter, the southwestem portion ruptured along the Nojima fault on the westem side of the Awajishima island along a strike slip motion with small vertical component. The other fault ruptured towards northeast beneath Kobe city with an almost pure strike slip movement. The seismic moment of this earthquake was determined to be 3x1026 by M.Kikuchi, Yokohama City Univ,. This value is relativdy small compared with the moment of the Northrigeg earthquake(M=1.0x1027) which occurred north of Los Angdes about a year ago. However, the January 17 quake caused us causalities of about 5,300 and extraordinary kinds of dameges like collapse of hundreds of houses and grave structural damage or break-down of elevated highways and Shinkansen tracks.

On the other hand, there were 9 permanent stations undertaking crustal movmat obsevations around the focal region when the earthquake Occurred(Figure 5). The Rokko-Tsurukabuto station is the dosest station to the epicenter, only 25 km away. Accordng to our preliminary analysis of the above crustal defomation data, no appereciable precursory change in Strain in all components of tiltmeters and extensometers were observed. Variations of ground-strains obseved in the area about 30 days before the earthquake observed at three stations are illustrated in Fig.6,7,8.

Several temporal observational points for seismicity, GPS, electromagnetic activity and geochemical variations were immediately installed near and around the epicentral area several days after the main shock. 25 temporal stations for seismic observations(Figure 9) were set up around the aftershock region to monitor transient seismic activitis in details?Hese observations have been carried out in cooperation With many universities Such as Hokaido,Hirosaki,Tohoku,Tokyo,Nagoya,Kyushu and Kagoshima university.

For detectings post-seismic deformations and pre-sliP of large aftershock, GPS Consortium of Japnese University set up 23 temporary stations on Feb.19-22 just 2 days after the present mainshock along or close to the fault in Kobe and Awajishima island as shown in Figure 10. In four stations indicated by open circles, real-time kinematic GPS observations sampled at 1 second interval have been conducted to monitors the real-time motions between sites across the fault(Figure 11). In the Figure, from the top to the bottom, are shown the latitude, the longitude, the height chage of a rover site relative to a base site in meter, and the number of GPS satellites and PDOP. All seismicity and GPS data have been telemetered to our center in DPRI, Uji.

However, there was no station installed to rcord electro-magnetic data around the epicenter due to the presence of relatively high artificial noise in Kobe City. Approximately one week after the main shock, we set up 9 temporal stations for the geomagnetic total foecd observation at Awajishima island and in the northem side of Rokko mountain to detect the geomagnetic total force change due to streess releass in the upper crust. These observation points were installd in cooperration with Tokyo Institute of Technology University, Earthquake Research Institute of Tokyo University and Dapartment of Science of Kyoto University. 6 stations were also set up in the same arre to measure the natural earth's potential. The location of the temporal stations accompanid with some partial results is shown in Fig.12,Fig,13 and Fig,14. No remarkable change is observed up to the present using these observation points.

Temporal variations in water contents and temperature at 3 hot or mineral springs in Hyogo and Tottori prefectures before and after the earthquake were examined. Location of observation stations are shown in Fig.15. Preliminary analysis shows that there are remarkable co-seismic changes in water tmperature and electrical conductivity were very clear at YDN and YSO, respectively. It is important to point that the flow of the groundwater at CSW stoppd co-seismically. A preliminary result is shown in Fig.16.

to bee continue