Deformation mechanisms and fluid behavior in a shallow, brittle fault zone during coseismic and interseismic periods: Results from drill core penetrating the Nojima Fault, Japan

Abstract

This paper describes the results of petrographical and meso-to microstructural observations of brittle fault rocks in cores obtained by drilling through the Nojima Fault at a drilling depth of 389.52 m. The zonation of deformation and alteration in the central zone of the fault is clearly seen in cores of granite from the hanging wall, in the following order: (i) host rock, which is characterized by some intragranular microcracks and in situ alteration of mafic minerals and feldspars; (ii) weakly deformed and altered rocks, which are characterized by transgranular cracks and the dissolution of mafic minerals, and by the precipitation of zeolites and iron hydroxide materials; (iii) random fabric fault breccia, which is characterized by fragmentation, by anastomosing networks of transgranular cracks, and by the precipitation of zeolites and iron hydroxide materials; and (iv) fault gouge, which is characterized by the precipitation of smectite and localized cataclastic flow. This zonation implies that the fault has been weakened gradually by fluid-related fracturing over time. In the footwall, a gouge layer measuring only 15 mm thick is present just below the surface of the Nojima Fault. These observations are the basis for a model of fluid behavior along the Nojima Fault. The model invokes the percolation of meteoric fluids through cracks in the hanging wall fault zone during interseismic periods, resulting in chemical reactions in the fault gouge layer to form smectite. The low permeability clayrich gouge layer sealed the footwall. The fault gouge was brecciated during coseismic or postseismic periods, breaking the seal and allowing fluids to readily flow into the footwall, thus causing a slight alteration. Chemical reactions between fluids and the fault breccia and gouge generated new fault gouge, which resealed the footwall, resulting in a low fluid condition in the footwall during interseismic periods.

Architectural evolution of the Nojima fault and identification of the activated slip layer by Kobe earthquake

“…[3] Several fault zone drilling project had been carried out since 1996 for the first fault zone penetration project after the 1995 Kobe earthquake [Ando, 2001]. Since then, there are have been three Nojima drilling projects from the years 1996 to 1999 [Lin et al, 2001;Tanaka et al, 2001aTanaka et al, , 2001bTanaka et al, , 2001cOhtani et al, 2001;Kobayashi et al, 2001], the Taiwan Chelungpu Fault zone drilling for shallow ones (2000 -2001) Heermance et al, 2003] and for the deeper one (2004 -2006) after the 1999 Chi-Chi, Taiwan earthquake, and the Atotsugawa Fault tunneling project (1998 -2003) [Forster et al, 2002;Tanaka et al, 2007]. In view of the completeness of the three Nojima fault drillings after almost 10 years of the 1995 Kobe earthquake, we would like to present a quantitative estimate of evolution of fault zone architecture, its relation to the primal slip surfaces activated by the 1995 Kobe earthquake, and petrological/chemical characteristics of the slip surface.…”

“…The reason for using NIED core is because it contains almost entire fault rocks across the fault zones including a possible slip layer of Kobe earthquake [Ikeda, 2001;Kobayashi et al, 2001;Tanaka et al, 2001a;Matsuda et al, 2004]. For other Nojima cores (GSJ and DPRI cores), slip layer could be plausibly assumed [Tanaka et al, 2001b[Tanaka et al, , 2001cFukuchi and Imai, 2001;Lin et al, 2001] but could be ambiguously defined. Also, petrological and geochemical examinations performed for the NIED core and geophysical logs from NIED borehole are all available for this study.…”