A large hydrothermal reservoir beneath Taal Volcano (Philippines) revealed by magnetotelluric resistivity survey: 2D resistivity modeling

Taal Volcano, located in the southwestern part of Luzon Island, Philippines, has frequently experienced catastrophic eruptions from both the Main Crater on Volcano Island and flank eruptions. These eruptions have been magmatic, phreatomagmatic, and hydrothermal, with the latter implying the existence of a large-scale hydrothermal system beneath the volcano. We conducted an electrical resistivity survey using the magnetotelluric method in order to identify the location and geometry of the hydrothermal reservoir and sealing cap rock. Two-dimensional inversion using the observed data indicates four similar resistivity sections. The structure at shallow depths corresponds to volcanic deposits and an aquifer. Below 1 km, the structure features a relatively resistive zone beneath the main crater surrounded by a conductive shell. We interpreted these to be a large hydrothermal reservoir with an impermeable cap rock sealing it. Recent ground deformation detected by GPS measurements suggests that the hydrothermal reservoir is active. The interpreted cap rock thins just beneath the main crater and could easily be destroyed by an imbalance in the hydrothermal system. We conclude that this hydrothermal reservoir plays a significant role in driving catastrophic eruptions that begin with a hydrothermal explosion at the main crater.     

Numerical modelling of the three-dimensional stress field in southwestern Japan

It has been inferred from the focal mechanism of earthquakes and their hypocenter distribution (Shiono, 1977) that the stress field in southwestern Japan indicates complicated features; a NW-SE compression at shallow depths along the Nankai trough, an E-W or ESE-WNW compression in the inland crust, an extension parallel to the leading edge of the Philippine Sea plate at subcrustal depths in the region from the southern Chubu to northwestern Shikoku, and a down-dip tension beneath the Kyushu island.In order to investigate possible sources of these complex features of the stress state, a three-dimensional finite element method is employed to model the configuration of the subducting Philippine Sea plate, taking into consideration the following three possible types of forces:

1. (1) A negative buoyancy due to the density contrast between the subducting plate and the surrounding mantle.

2. (2) A northwestward compressive force generated by the movement of the Philippine Sea plate.

3. (3) A westward compressive force due to the movement of the Pacific plate.

For various combinations of different magnitudes of these forces, and of different elastic moduli, the stresses at a number of selected sites are calculated, and their directions are compared with those inferred from the focal mechanism and other geophysical information.It is found that the observed extensional stresses parallel to the leading edge of the subducting Philippine Sea plate may be caused mainly by the negative buoyancy. The northwestward compressive force may not play an important role in generating the complex stress field in southwestern Japan. The observed E-W compressional stress field prevailing in the inland region appears to result mainly from the westward-moving Pacific plate. The present results suggest that if a thin low-velocity transitional layer exists just above the subducting Philippine Sea plate, it could give appreciable effects on the mechanism of low-angled thrust faulting off the Kii peninsula and the Shikoku island.The magnitude of the shear stress in the continental crust and in the subducting plate is estimated to be of the order of several hundred bars, although the calculated shear stresses are considerably affected by the configuration of the subducting plate and also by the applied forces.It is interesting that the stress distribution appears to have some relations to seismicity of subcrustal earthquakes, and to the rupture process of large thrust earthquakes along the plate boundary.     

Estrogenic activity in sediments contaminated by nonylphenol in Tokyo Bay (Japan) evaluated by vitellogenin induction in male mummichogs (Fundulus heteroclitus)

Estrogenic activity was determined in sediments collected from Tokyo Bay. Sampling was performed at five stations including the site near the sewage treatment plant. The most estrogenic sediment collected near the sewage treatment plant was fractionated into ten fractions using normal-phase high-performance liquid chromatography. Chemical analysis was carried out for each fraction and nonylphenol (NP, 20,700 ng g⁻¹ dry wt) was detected at a higher concentration than estron (2.39 ng g⁻¹ dry wt) and 17β-estradiol (<0.7 ng g⁻¹ dry wt). Furthermore, each fraction was administered to male mummichogs (Fundulus heteroclitus), and vitellogenin (Vtg) was measured after two weeks. The induction of Vtg was observed; this estrogenic potency could be attributed to the NP content in this fraction. This is the first report to suggest that the high NP concentration in the sediments from Tokyo Bay has the potential to induce Vtg in wild fish.     

Simulation of the Transition of Earthquake Rupture from Quasi-static Growth to Dynamic Propagation

-- We have simulated a rupture transition from quasi-static growth to dynamic propagation using the boundary integral equation method. In order to make a physically reasonable model of earthquake cycle, we have to evaluate the dynamic rupture propagation in the context of quasi-static simulation. We used a snapshot of the stress distribution just before the earthquake in the quasi-static simulation. The resultant stress will be fed back to the quasi-static simulation. Since the quasi-static simulation used the slip-and time-dependent constitutive relation, the friction law itself evolves with time. Thus, we used the slip-weakening constitutive relation for dynamic rupture propagation consistent with that used for the quasi-static simulation. We modeled a San Andreas type strike-slip fault, in which two different size asperities existed.     

Detection of crustal inhomogeneity in the Nojima Fault zone using earth current noise

Abstract A resistivity survey method using artificial telluric noise was examined and applied to a field of a fault zone. The electric earth current was measured at 50 sites in the Nojima Fault zone, which is in the northwestern part of Awaji Island, southwestern Japan. The dominant component of the observed electric field is supposed to be leakage currents from DC electric railways running outside the island. Amplitude and polarization of the stray current were systematically investigated and were revealed to represent the subsurface electrical structure of the study area. Some features on the fault zone's electrical structure have been pointed out, including: (i) an electrical boundary that corresponds to a geological one between granite (resistive) and sediments (conductive); and (ii) a low resistivity spot on the surface rupture of the earthquake fault. The structure estimated in the present study is both qualitatively and quantitatively consistent with previous resistivity surveys done using other methods pursued in the same area. It shows the validity of the 'stray current method' as one that is easy and uses low-cost resistivity exploration tools in a region where the effect of artificial noise caused mainly by leakage currents from electrical railways cannot be ignored.     

3-D Simulation of Tectonic Loading at Convergent Plate Boundary Zones: Internal Stress Fields in Northeast Japan

The evidence of east-west compression in northeast Japan has been reported by many investigators on the basis of geodetic, geologic and geomorphic data, but its origin still remains far from understood. In the present study we have proposed a mechanical model of tectonic loading at convergent plate boundary zones, and demonstrated its validity through the numerical simulation of internal stress fields in northeast Japan with realistic 3-D geometry of plate interfaces. At convergent plate boundary zones, in general, a part of plate convergence is consumed by steady slip along plate interfaces, and the remaining part by inelastic deformation (seismic faulting, aseismic faulting, and active folding) of overriding plates. Such a plate boundary process to be called ``partial collision' can be quantitatively described by introducing a collision rate defined as c = 1 − steady slip rate at plate interfaces/plate convergence rate. By this definition, we can simply represent the mechanical process of partial collision, which includes total subduction (c = 0) and total collision (c = 1) as two extreme cases, in terms of steady slip rates at plate interfaces. On the basis of elastic dislocation theory, first, we numerically computed the internal stress fields in northeast Japan produced by the total subduction of the Pacific plate beneath the North American plate, however the computed stress pattern was opposite in sense to observations. Then, we computed the internal stress fields by taking c = 0.1 on average, and succeeded in reproducing the observed east-west compression in northeast Japan. This indicates that the concept of partial collision is essential to understand the mechanism of intraplate tectonic loading.     

Quasi-Static Strain and Stress Fields due to a Moment Tensor in Elastic–Viscoelastic Layered Half-Space

We derived explicit expressions in the time domain for 3-D quasi-static strain and stress fields, due to a point moment tensor source in an elastic surface layer overlying viscoelastic half-space under gravity. The expressions of strain in the elastic surface layer were directly obtained from the expressions of displacement in our previous paper. The conversion of strain into stress is easy, because the stress–strain relation of elastic material is linear. In the viscoelastic substratum, the expressions of strain were obtained by applying the correspondence principle of linear viscoelasticity to the associated elastic solution. The strain–stress conversion is not straightforward, as the stress–strain relation of viscoelastic material is usually given in a differential form. To convert strain into stress, we used an integral form of the stress–strain relation instead of the usual differential form. The expressions give the responses of elastic half-space at \( t = 0 \) , and the responses of an elastic plate floating on non-viscous liquid at \( t = \infty \) . The moment tensor is rationally decomposed into the three independent force systems, corresponding to isotropic expansion, shear faulting and crack opening, and so the expressions include the strain and stress fields for these force systems as special cases. As the first numerical example, we computed the temporal changes in strain and stress fields after the sudden opening of an infinitely long vertical crack cutting the elastic surface layer. Here, we observe that the stress changes caused by the sudden crack opening gradually decay with time and vanish at \( t = \infty \) everywhere. After the completion of stress relaxation, a characteristic pattern of shear strain remains in the viscoelastic substratum. Since the strain and stress fields at \( t = \infty \) can be read as the strain- and stress-rate fields caused by steady crack opening, respectively, this numerical example demonstrates the realization of a steady stress state supported by steady viscous flow in the asthenosphere, associated with steady seafloor spreading at mid-ocean ridges. For the second numerical example, we computed the temporal changes in strain and stress fields after the 2011 Tohoku-oki mega-thrust earthquake, which occurred at the North American-Pacific plate interface. In this numerical example, the stress changes caused by coseismic fault slip vanish at \( t = \infty \) in the viscoelastic substratum, but remain in the elastic surface layer. The coseismic stress changes (and also strain changes) in the elastic surface layer diffuse away from the source region with time, due to gradual stress relaxation in the viscoelastic substratum.     

A heating process of Kuchi-erabu-jima volcano,Japan, as inferred from geomagnetic field variations and electrical structure

Since August 2000, we have recorded the total intensity of the geomagnetic field at the summit area of Kuchi-erabu-jima volcano, where phreatic eruptions have repeatedly occurred. A time series analysis has shown that the variations in the geomagnetic field since 2001 have a strong relationship to an increase in volcanic activity. These variations indicate thermal demagnetization of the subsurface around the presently active crater. The demagnetization source for the early variations, until summer 2002, was estimated at about 200 m below sea level. For the variations since 2003, the source was modeled on the basis of the expansion of a uniformly magnetized ellipsoid. The modeling result showed that the source is located at 300 m above sea level beneath the crater. We carried out an audio-frequency magnetotelluric survey with the aim of obtaining a relation between the demagnetization source and the shallow structure of the volcano. A two-dimensional inversion applied to the data detected two good conductors, a shallow thin one which is restricted to a region around the summit area, while the other extends over the edifice at depths between 200 and 800 m. These conductors are regarded as clay-rich layers with low permeability, which were assumed to be generated through hydrothermal alteration. The demagnetization source for the early variations was possibly located at the lower part of the deep conductor and the source after 2003 lies between the two conductors, where groundwater is considered to be abundant. Based on these results, as well as on seismological, geodetic, and geochemical information, we propose a heating process of the Kuchi-erabu-jima volcano. In the initial stage, high-temperature volcanic gases supplied from the deep-seated magma remained temporarily at the level around the lower part of the less permeable deep conductor since the ascent path had not yet been established. Then, when the pathway developed as a result of repeated earthquakes, it became possible for a massive flux of volcanic gases to ascend through the conductor. The high temperature gases reached the aquifer located above the conductor and the heat was efficiently transported to the surrounding rocks through the groundwater. As a consequence, an abrupt increase of the gas flux and diffusion of the heat through the aquifer occurred and the high-temperature zone expanded. Since the high-temperature zone is located beneath another conductor, which acts as caprock, we assume that the energy of the phreatic explosion is accumulated there.     

Isophot Mapping Observations of Carbon Stars

Mapping observations have been made toward five carbon stars in the far-infrared using ISOPHOT, an imaging photo-polarimeter on board the Infrared Space Observatory. Cold, very extended dust shells are clearly revealed in two of them. Y CVn is surrounded by a very extended, detached dust shell, which indicates a sudden decline of the mass-loss by two orders of magnitude in the last (1-2) × 104 years on a short time scale. The Hipparcos parallax reinforces our previous conclusion that Y CVn is a J-type carbon star on the asymptotic giant branch. U Ant shows a double shell structure, a compact dust shell surrounded by a very extended one. The outer shell has a brightness comparable to the dust shell of Y CVn. The structure indicates that there were two different high mass-loss phases separated by a period with a much lower mass-loss rate in between the two. The structure is consistent with the double dust shell proposed for this star by Izumiura et al. (1997) based on a detailed investigation of IRAS survey data. This revised version was published online in August 2006 with corrections to the Cover Date.     

Electrokinetic phenomena associated with a water injection experiment at the Nojima fault on Awaji Island, Japan

Abstract Self-potential variations were measured to estimate the magnitude of electrokinetic and hydrological parameters (zeta potential and permeability) of the Nojima Fault zone in Awaji, Japan. The study observed self-potential variations that seemed to be associated with water flow from the injection well to the fracture zone, which were induced by turning the injection on and off. Amplitudes of the variations were a few to 0.03 V across 320–450 m dipoles. These variations can be explained well with an electrokinetic model. The quantity k/ζ (permeability/zeta potential) is in the range 1.6 × 10⁻¹³− 5.4 × 10⁻¹³ m²/V. Permeability of the Nojima fault zone can be estimated as approximately 10⁻¹⁶–10⁻¹⁵ m² on the assumption that the zeta potential is in the range –0.01 to –0.001 V.