Kenji Kanjo

A water well sensitive to seismic waves

The water level in an observation well (EDY) has been monitored with high frequency sampling (1/sec). Large changes in the water level are observed in this well, which correspond to dynamic strain changes due to the passage of long period seismic waves. For example, fluctuations of more than 200 cm were observed after a M 8.1 earthquake, 1200 km from the observation site. The response characteristic of the water level to strain change was investigated by comparison with nearby volumetric strain changes: the response curve has a peak of sensitivity and a phase transition at periods of 18∼19 seconds. At the resonant period, the sensitivity is as large as 15 mm per 10−9 strain.

Seismic evidence for a thinner mantle transition zone beneath the South Pacific Superswell

Broadband seismograms recorded by the seismic stations deployed on oceanic islands in the South Pacific for two deep earthquakes in 1998 are used to investigate the mantle transition zone structure beneath the South Pacific, where a large-scale hot plume might ascend from the core-mantle boundary (CMB). By stacking S waves reflected at the mantle transition zone discontinuities, we find clear signals associated with the 410-km and 660-km discontinuities. The transition zone thickness, which is constrained from the travel time difference between the reflected waves from the 410-km and 660-km discontinuities, is observed to be approximately 15 km thinner beneath the South Pacific Superswell than that beneath other regions. The result suggests that the transition zone temperature beneath French Polynesia is approximately 100∼200 K higher than the surrounding mantle.

Response of a volcanic conduit to step‐like change in magma pressure

Freely damped oscillations associated with strain offsets were observed by a nearby volumetric strainmeter during the 1986 Izu-Oshima Volcano eruption, Japan. These events are interpreted as a response of the magma-filled conduit to step-like changes in magma pressure. There were dramatic changes in the characteristic frequency of oscillation and the polarity of initial motion before and after the 6-hour interval of the enormous ground deformation due presumably to magma intrusion. The frequency changed from 42.6 to 21.3 mHz and the polarity changed from contraction to expansion. We propose a conceptual conduit model to explain this polarity reversal, where magma is episodically supplied to the conduit before but is episodically drained back to the reservoir after an extensive magma escape away from the reservoir as dike intrusion. Stress caused by the dike intrusion is large enough to break the lid of the conduit, thus to change a boundary condition. This change in boundary condition reduces the frequency of oscillation by a factor of 2. In this model longitudinal wave speed of the magma fluid is 340 m/s if the conduit length is taken to be 4 km.

Strain offsets with monotonous damped oscillations during the 1986 Izu‐Oshima volcano eruption

The sequential process of the 1986 Izu-Oshima volcano eruption, Japan, was recorded by a volumetric strain meter near the erupted summit. The records show occasional strain offsets accompanying long-period damped oscillations, which we interpret as responses of the volcanic system to perturbations in magmatic pressure. The polarities of strain offsets were contractive for all events before the enormous ground deformation of November 21 due presumably to the intensive dike intrusion. On the contrary, all events after that had expansive polarities. The characteristic frequencies of oscillations were ∼42.6 mHz for the events before, changing abruptly to half that after the ground deformation. In our model, which is a one-dimensional magma-filled conduit, the boundary condition at the top of the conduit changed from a closed to an open end upon a void formation at the top of the magma head due to its descent immediately after the dike intrusion. Before the magma escape from the reservoir as a dike, magma was episodically supplied from the reservoir to the conduit, and thus contraction events were generated. After the magma escape, magma left in the conduit was episodically drained back to the reservoir to generate expansion events. This episodic magma supply or drainage was modeled by a prescribed pressure change at the bottom of the conduit. We solved the free oscillation problem of the conduit pipe with an elastic sidewall to calculate the resultant volumetric strain outside the pipe. Physical properties of the magma and the excitation mechanism are inferred from the application of our formulation to the observation. The conduit length and fluid acoustic velocity are estimated to be ∼10 km and 430 m/s, respectively, suggesting that the fluid is composed of a gas-liquid mixture. If the conduit radius is taken to be 40 m and if the damping of oscillation occurred by viscous dissipation, the fluid viscosity is 108–109 Pa s, a reasonable estimate for the basaltic magma. The applied pressure changes at the bottom of the conduit are in a range from 105 to 106 Pa with risetimes of the order of 4 to 100 s.

Application of Mwp to the Great December 26, 2004 Sumatra Earthquake

A great earthquake of Mw9.0 (Harvard) occurred off of northwestern Sumatra on December 26, 2004 (UTC), causing an unprecedented tsunami disaster. An earthquake of Mw8.6 (Harvard) then occurred on March 28, 2005 (UTC), about 160 km to the southeast of the December event’s epicenter. The Matsushiro Seismological Observatory of Japan Meteorological Agency determined magnitudes of M8.8 and M8.7 respectively for these events using the Global Seismic Network’s (GSN) Live Internet Seismic Server (LISS) data. The West Coast/Alaska Tsunami Warning Center (WC/ATWC), and the Pacific Tsunami Warning Center (PTWC) both use Mwp to rapidly evaluate moment magnitude. The WC/ATWC calculated a mean Mwp of 8.0 for the multiple event on Decmber 26, 2004. Using data from the IRIS station MAJO, we determined an Mwp of 8.5 by using a distance-dependent apparent P-wave velocity (α = 0.16Δ + 7.9 km/sec.) instead of a constant apparent P wave velocity (α = 7.9 km/sec), for α in the original equation for Mwp. The corrected Mwp value of 8.5 is much closer to the total moment magnitude of the multiple ruptures of the complex December 26 main-shock, and is useful as a first magnitude estimation to evaluate possible tsunami generation.

Detection of shallowest submarine seismicity by acoustic coupled shear waves

An outstanding slow wave train similar to T wave in appearance has been observed on records of an ocean bottom seismometer array, off Boso, central Japan. These slow wave trains are not observed on land, are usually spindle-shaped and dominant in vertical component, and have apparent velocities of ∼1.3 km/s along the array. In order to understand the observations, we calculate synthetic seismograms by a finite difference method for horizontally stratified layers of seawater, sediment, and basement, which show a well-defined wave train consistent with the observations only if the source is placed above the basement layer. This result and normal mode calculations indicate that the slow wave train comprises essentially the sediment-trapped shear waves coupled with ocean acoustic waves. The observed events of slow wave trains occur near the boundary between the outer arc high and the forearc basin, where vertical differential motion has been suggested to occur. We suggest that these small events represent the shallowest crustal activity unique to the triple junction off Boso.