Effects of fluid pressures to the seismic velocity of crustal rocks

Introduction Water of the earth interior is mainly supplied at the subduction zone and has important role on seismic activity and volcanism in island arc. It is suggested that slow slip events and tremors occurring at this region are related to water (e.g., Obara, 2000). Based on the seismic tomography beneath Kanto district, high Poisson's ratio area (~0.337) was observed and suggested weak seismic coupling at plate interface (Kamiya and Kobayashi, 2000). Similar high Poisson's ratio is detected beneath Tonankai and Shikoku district, exceeding 0.3. Those regions correspond to the plate boundary generating slow slip events or tremors (Kodaira et al., 2004 ; Shelly et al., 2006). Because relatively young oceanic plates are subducting in districts from Kanto to Sikoku, antigorite which Poisson's ratio is ~0.29 may exist stably in those areas. In this case, the observed high Poisson's ratio requires excess pore fluids in addition to the serpentinized mantle. In order to clarify geometry and the abundance of water, we investigate seismic velocity of crustal rocks under high confining pressure and pore fluid pressure. Experimental methods For the measurement of seismic velocity, we used the hydraulic pressure vessel in Hiroshima University, in which seismic velocity was calculated by using pulse echo method. We used the Aji granite as a test sample, which was prepared into a cylindrical shape with 20 mm diameter and 5-10 mm length. The top and bottom of sample were polished mirror surfaces within 0.001 mm difference. We measured seismic velocity under dry and wet conditions. In the later case, distilled water is supplied into the sample with pore pressure up to 100 MPa, and we also measured permeable time of water at Pc = 20 MPa, Pp = 0 MPa and Pc =40 MPa, Pp =20 MPa. Results and discussion Under dry experiments, seismic velocities of granite were measured up to confining pressure as high as 200 MPa. Calculated velocities were Vp = 5.900 km/s, Vs = 3.478 km/s at pressure of 200. These values are lower than Christensen (1996)'s experiments (Vp = 6.246 km/s, Vs = 3.669 km/s), but from the measurement of both compression and decompression process, it is confirmed that velocity variation has the reproducibility and there is an effect depending on cracks or pores closed by high confining pressure. Under wet experiments, at first, permeable time of water was measured. We found that water completely percolates into the sample around eight hours. Next, velocities of granite were measured up to confining pressure and pore pressure as high as 120 and 100 MPa (effecting pressure of 20 MPa). At confining pressure of 20 MPa before raising pore pressure, velocities were Vp = 5.414 km/s and Vs = 3.091 km/s. After confining pressure and pore pressure increased to 40 and 20 MPa, Vp was significantly increased to 5.869 km/s, while Vs was slightly changed to 3.124 km/s. Poisson's ratio was increased from 0.258 to 0.302. After the pore fluid penetrates homogeneously into the samples, both velocities were little change and became nearly constant values. These variations are similar to previous experiments (e.g., Saito, 1987). However, under high pore pressure, signal from the sample reflection is very weak and therefore these values may have large uncertainty. We try to fix this issue and hope to present the effect of pore pressure on seismic velocity in the coming AGU meeting.