In situ stress measurements in NIED boreholes in and around the fault zone near the 1995 Hyogo‐ken Nanbu earthquake, Japan

Abstract

The 1995 Hyogo-ken Nanbu (Kobe) earthquake, M7.2, occurred along the northeast-south-west trending Rokko-Awaji Fault system. Three boreholes of 1001 m, 1313 m and 1838 m deep were drilled in the vicinity of the epicenter of the earthquake. Each borehole is located at characteristic sites in relation to active faults and the aftershock distribution. In particular, the Nojima-Hirabayashi borehole [Hirabayashi National Research Institute for Earth Science and Disaster Prevention (NIED) drilling] in Awaji Island was drilled to a depth of 1838 m, approximately 320 m southeast from the surface rupture of the Nojima Fault, and it crosses fracture zones below a depth of 1140 m. In situ stress measurements by the hydraulic fracturing method were conducted in these boreholes within 1.5 years after the earthquake. Measurement results suggest the following: (i) Differential stress values are very small, approximately 10 MPa at a depth of 1000 m at each site; (ii) the orientation of maximum horizontal compression is almost the same in the boreholes, perpendicular to the surface trace of the faults, north-west-south-east; (iii) fault types estimated from the state of stress differ among these sites; and (iv) the differential stress value just beneath the fault fracture zone decreases abruptly to one-half of that above the fault zone in the Hirabayashi NIED drilling. These features support the idea that the shear stress along the Rokko-Awaji Fault system decreased to a low level just after the earthquake.Key words: 1995 Hyogo-ken Nanbu earthquake, active fault, hydraulic fracturing, in situ stress, Kobe earthquake, Nojima Fault, scientific drilling. the materials, the present state of the stress and strain, and the fluid pressure and permeability in deeper fault fracture zones in order to better understand fault dynamics. With this in mind and in considering the results obtained from the other drill holes that are summarized herein, we at the National Research Institute for Earth Science and Disaster Prevention (NIED), drilled three boreholes in the vicinity of the epicenter of the 1995 Hyogo-ken Nanbu (Kobe) earthquake (M7.2). One of these boreholes, Nojima-Hirabayashi on Awaji Island (Hirabayashi NIED drilling), was drilled to a depth of 1838 m and directly intersected the Nojima Fault that appeared as a surface scarp during the earthquake. The main purpose of these drillings was to quantitatively evaluate the stress states, permeability, and heat flow in the fault

Recovery of Stress During the Interseismic Period Around the Seismogenic Fault of the 1995 Mw 6.9 Kobe Earthquake, Japan

“…Based on the orientations of the breakouts, the σ Hmax azimuth of the group (i) varied with depth; the average σ Hmax azimuths of the upper and lower parts are estimated to be 176° and 138°, respectively (Figure a). This average σ Hmax azimuth in lower part is approximately perpendicular to the general trend of the Nojima Fault, similar to that estimated from in situ measurements carried out in 1997, for example, two years after the Kobe earthquake (Figure ; Ikeda et al, ; Tsukahara et al, ). In contrast, the average σ Hmax azimuth of group (iii) is oriented 241°, subparallel to the current regional compressive stress direction of E–W (Katao et al, ; Figure b).…”

Reconstruction of seismic faulting by high‐velocity friction experiments: An example of the 1995 Kobe earthquake

“…The friction coefficient of the Nojima fault was estimated to decrease from 0.6 to less than 0.3 during the Kobe earthquake [ Ikeda et al , 2001; Tadokoro et al , 2001; Tsukahara et al , 2001; Yamamoto and Yabe , 2001; Yamashita et al , 2004]. The high‐velocity friction experiments on the Nojima fault gouge show that the frictional strength decreased from 0.63 to 0.18.…”

“…The Nojima fault, southwest Japan, is one of the faults that moved during the Kobe earthquake [ Nakata and Yomogida , 1995]. In‐situ stress measurements at boreholes drilled along the fault after the earthquake and the deformation rate analysis (DRA) of the core samples indicated μ r of less than 0.3 [ Ikeda et al , 2001; Tadokoro et al , 2001; Tsukahara et al , 2001; Yamamoto and Yabe , 2001]. μ i was estimated to be about 0.6 from the difference of stress orientation before and after the earthquake [ Yamashita et al , 2004].…”

Frictional heat from faulting of the 1999 Chi‐Chi, Taiwan earthquake

“…Again, the values of μ are doubled if the maximum estimate of thermal anomaly is taken. For reverse fault regimes, the horizontal compressive stress is reported to be 2 to 4 times the vertical stress [ Sibson , 1974; Brace and Kohlstedt , 1980; Ikeda et al , 2001]. For this range of values, the level of friction needed to produce the thermal anomaly is quite low, with μ estimated to be 0.04 to 0.06 using a fault dip of 30°.…”