Stress field change around the Mount Fuji volcano magma system caused by the Tohoku megathrust earthquake, Japan

Abstract

Crustal deformation by the M _w 9.0 megathrust Tohoku earthquake causes the extension over a wide region of the Japanese mainland. In addition, a triggered M _w 5.9 East Shizuoka earthquake on March 15 occurred beneath the south flank, just above the magma system of Mount Fuji. To access whether these earthquakes might trigger the eruption, we calculated the stress and pressure changes below Mount Fuji. Among the three plausible mechanisms of earthquake–volcano interactions, we calculate the static stress change around volcano using finite element method, based on the seismic fault models of Tohoku and East Shizuoka earthquakes. Both Japanese mainland and Mount Fuji region are modeled by seismic tomography result, and the topographic effect is also included. The differential stress given to Mount Fuji magma reservoir, which is assumed to be located to be in the hypocentral area of deep long period earthquakes at the depth of 15 km, is estimated to be the order of about 0.001–0.01 and 0.1–1 MPa at the boundary region between magma reservoir and surrounding medium. This pressure change is about 0.2 % of the lithostatic pressure (367.5 MPa at 15 km depth), but is enough to trigger an eruptions in case the magma is ready to erupt. For Mount Fuji, there is no evidence so far that these earthquakes and crustal deformations did reactivate the volcano, considering the seismicity of deep long period earthquakes.

Appendices

Appendix: sensitivity of simulation results

Our numerical simulation using FEM depends on the assumptions underlying the model. In our analysis, we use seismic fault parameters derived from the inversion of the observation data and the physical properties of the crust and magma are based on the results of seismic tomography. This appendix briefly discusses the sensitivity of our numerical simulation results with respect to three points: (1) the heterogeneity and topography of the crust, (2) the depth of magma reservoir, and (3) the fault model of Tohoku earthquake.

The heterogeneity and topography of the crust

Figure 9 presents examples of simulations of crustal deformation for a homogeneous medium with a no-topography model and for a heterogeneous medium with a topography model. The displacements and stress (in the figure we depict the trace of the stress tensor) suggest different profiles. For example, the peak of EW(x) displacement in the homogeneous-no-topography model is about 0.018 m, whereas that in the heterogeneous topography model is about 0.014 m. The estimated stress changes are on the same order, but the stress trace in the homogenous model has a peak of 0.16 MPa, while that in the heterogeneous model is about 0.06 MPa, less than half of the former case. Therefore, it is better to include the heterogeneity and the actual topography, though we can discuss the order of the stress changes in spite of the ambiguity of the crustal structure.

Fig. 9

Sensitivity of FEM numerically simulated crustal deformation on the heterogeneity and topography. The left plate illustrates the distribution of displacement and stress for a homogeneous medium with no topography; the right plates illustrate that for a heterogeneous medium with topography. In the figure, displacements 0, 1, and 2 correspond to the EW, NS, and UD components, and σ _trace is equal to the sum of σ _xx, σ _yy, and σ _zz

The depth of magma reservoir

The magma plumbing system beneath Mount Fuji is not clear but we assume it from hypocenter distribution and seismic tomography results. One of the most important factors is the depth of the magma reservoir, so we evaluate how the stress field depends on it. Figure 10 shows the EW cross section of σ _dif distribution by East Shizuoka earthquake beneath the summit (left) and the vertical distribution of stress components passing the center of magma reservoir (right). We assume two depths of magma reservoir as (a) 15 km and (b) 20 km. The red dashed circles correspond to the boundary of magma reservoir and it is noted that the σ _dif takes the maximum above the reservoir about (a) 1.35 MPa and (b) 1.28 MPa, respectively. Thus, the stress given to the magma system is effective to its depth.

Fig. 10

Differential stress σ _dif distributions for two cases of magma reservoir depths of a 15 km and b 20 km. The dashed red circles indicate the boundary of spherical magma reservoirs. The profile of stress components along the vertical axis is also shown, suggesting the maximum σ _dif is at above the magma reservoir

The fault model of Tohoku megathrust earthquake

For the evaluation of stress field change due to Tohoku earthquake, we applied the fault model by Ozawa et al. (2011), which is inverted by the GPS data and the displacements are constant on each of two major faults. Recent analysis by Simons et al. (2011) suggests detail slip distributions on the fault. Similarly, Koketsu et al. (2011) reported a fault model obtained by joint inversion of teleseismic, strong motion, geodetic and Tsunami data. In their model, the fault region is divided into 80 segments with individual displacement. As in Fig. 11, the Tohoku area has different displacement and stress distributions from those obtained as in Fig. 4 by Ozawa et al. (2011), but the stress distribution around Mount Fuji region has no significant difference, comparing Fig. 11 bottom with Fig. 5 top, since Mount Fuji is distant enough from Tohoku region.

Fig. 11

Disturbance by Tohoku earthquake calculated by the fault model of Koketsu et al. (2011), obtained by joint inversion of teleseismic, strong motion, geodetic and Tsunami data. The fault plane is divided into 80 segments with individual displacement. The disturbance around Mount Fuji is not significantly different from those by the fault model of Ozawa et al. (2011) in Figs. 3 and 4