In-situ Calibration of Borehole Strainmeter Using Green's Functions for Surface Point Load at a Depth of Deployment

Geological Survey of Japan, AIST established fourteen observation sites of borehole strain, tilt and groundwater at the Shikoku island, Kii peninsula and Tokai region for the prediction research of the Nankai and Tonankai earthquakes. This paper represents the results of in-situ calibration of the fourteen multi-component borehole strainmeters. For the in-situ calibration of the borehole strainmeter, amplitudes and phases of M₂ and O₁ tidal constituents are extracted from the borehole strainmeter data, and then the extracted tidal amplitudes and phases are compared with M₂ and O₁ constituents of theoretical tidal strain which are estimated by software packages such as GOTIC2 (Matsumoto et al, 2001). In these software packages, amplitudes and phases of theoretical M₂ and O₁ tidal strain are calculated from the solid Earth tide and oceanic tidal loading. These software packages usually use green’s function at surface observation for the surface point load to calculate tidal strain caused by the oceanic tidal loading. Kamigaichi (1998) showed significant difference between green’s function at depth = 0 and that at depth ≠q 0. Kamigaichi also made computer programs for the calculation of the green’s function at arbitrary depth for a surface point load, and he modified the GOTIC2 program to apply the green’s function at arbitrary depth to the calculation of theoretical strain caused by the oceanic tidal loading. We calculated theoretical tidal areal strain at ICU observation site whose distance from the coast is about 150m. We found that the theoretical tidal areal strain calculated by the modified GOTIC2 is about four times of that calculated by the original GOTIC2 at ICU, and the areal strain calculated by the modified GOTIC2 is consistent with the areal strain observed by borehole strainmeter. M₂ and O₁ constituents of the theoretical two-component tidal tilts calculated by the modified GOTIC2 are almost consistent with tidal tilts observed by borehole tiltmeter at ICU site. We extracted amplitudes and phases of M₂ and O₁ constituents from the multi-component borehole strainmeter data at fourteen observation sites, and applied calibration methods proposed by Gladwin and Hart (1985), Roeloffs (2010) and Hart et al. (1996). We selected the calibration method by Roeloffs (2010) for the nine of 14 multi-component borehole strainmeters and the isotropic calibration method by Gladwin and Hart (1985) for other five strainmeters, because residuals between theoretical values and observed values at the nine borehole strainmeters are significantly reduced when we apply the calibration method by Roeloffs (2010).