Usu volcano has erupted seven times since 1663. The last three eruptions occurred during the 20th century (1910, 1943 and 1977) and were observed by standard instruments. Although the three eruptions displayed different surface manifestations, they were, similarly, volcanisms derived from dacitic magmas.
n this paper, the three eruptions are compared mainly from the viewpoint of magma movements before and during the eruptions on the basis of geophysical data, i.e. explosive, seismic and thermal activities, ground deformations, and level changes in aquifers. Post-eruption activities are also comparatively reviewed.
Some phenomena related to the three eruptions are reinterpreted: for the 1910 eruption, the cryptodome model of the upheaval is rejected, and the tectonic structure of the craterlet line is proposed; for the 1943 and the 1977 eruptions, the ratios between seismic energy release and deformation are discussed; for the 1977 eruption, a tilt model of the sector deformation is proposed and relationships between anomalous changes in the aquifer level and ground upheavals are discussed.
Magnitudes of the three eruptions are estimated. Tentatively, the 1943 and 1977 eruptions are roughly of the same order of magnitude and the 1910 eruption is one order of magnitude smaller.

抄録全体を表示

This paper investigates the cause of annual variations in the geomagnetic total intensity that are often seen especially in the volcanic areas. As a hypothesis of the cause, a model was proposed, in which such a change is produced by changes in the inhomogeneous magnetization of near-surface rocks due to temporal changes of the atmospheric temperature. This hypothesis was tested by field and laboratory experiments. First, amplitude and phase difference of annual variations in the total intensity and ground temperature data were determined by time series analyses. Considering thermal diffusion from the surface into the ground, the phase difference between the total intensity and temperature was converted to a characteristic depth, and then the amplitude of annual temperature variation at the depth was estimated. Finally, the observed total intensity variations were compared with the expected change on the basis of the temperature dependence of rock's magnetization obtained by a laboratory experiment and the local magnetic anomaly obtained by a magnetic survey at each magnetometer site. A good agreement between the observed and expected changes was obtained, which strongly suggests that the hypothesis is correct. It was also shown that a correction of annual variations by using temperature data will enable us more accurate detection of volcanomagnetic signals.

抄録全体を表示

Regular measurements of [OI] 630 nm nightglow have been carried out at Kolhapur (16.8°N, 74.2°E, dip lat 10.6°N) in India with the help of tilting filter photometers during clear moonless nights. Several characteristic nightglow data sets are presented in the report for geomagnetically quiet and disturbed nights. A comparison has been made between the simultaneous nightglow and the ionospheric parameters of virtual F-layer height and critical frequency of the F-layer at Ahmedabad (23.02°N, 72.6°E). In general, a good correlation (c.c.= 0.8 to 0.9) is found between the measured airglow emission and F-region peak electron densities during quiet periods but for intense magnetic disturbances the airglow fluctuations are mainly controlled by sharp changes in height variations of the h′ F layer. The period of airglow variation generally corresponds to periods of F-region height variation during magnetic disturbances. Using a long series of data between December 1989 and April 1996, the mean seasonal nocturnal variations of 630 nm emission have been compared during high and low solar activity periods.

抄録全体を表示

Combining the data of in-situ measurements of ionospheric current, J_m by rocket-borne instruments and the ground based geomagnetic H field close to the magnetic equator a linear relation has been found between the peak current density J_m and the daily range of H, (R_H). This relationship has been used to convert long series of R_H data into J_m. Combining J_m and the E-region peak electron density N_m, the electron velocity in the ionosphere, V_E has been calculated. It is shown that after all corrections are made of the solar zenith angle variations, the ionospheric current as well as electron drift in American and Indian sectors show strong equinoctial maxima, the mean values of both the parameters are larger at American than in Indian sector. The solar cycle variation of the electrojet current is primarily due to the variations of N_mE, and not due to the variations of electric field. The diurnal variation of the electric field with peak at 09-10 LT interacting with noon peak of N_mE making ΔH to peak at an hour earlier than noon. It is stressed to realise the importance of electric field in diurnal, seasonal and longitudinal variations of the equatorial electrojet current.

抄録全体を表示

A study is carried out on the mean monthly values of in situ observations of solar wind velocity (V) and the intensity of interplanetary magnetic field, B to elucidate their long term variations using the technique of singular spectrum analysis. It is shown that B exhibits a clear solar cycle signal with progressively deepening minimum and a well-defined longer period variation but V is marked by a -9-yr cycle. Time variations in the amplitude of 16-month periodicity, observed sporadically in the velocity earlier by others are clearly brought out, as also the characteristic amplitude changes in a 42-month signal in velocity. It appears that time intervals which show certain significant fluctuations in B are marked by the absence of similar signal in V.
aa index is next used as a proxy for solar wind velocity, after establishing a good correlation between Ap and observed V so that we could examine the evolution of different periodicities over 14 solar cycles. A significant trend with least value in 1900 and a near-linear rise up to 1960 is shown to be the main feature of the velocity change. The solar cycle component in V lags the solar activity peak by -22 months. Streams emanating from coronal holes in the declining phase seem to be the most dominant contributor to the 11-year variation in velocity. The anomalous pattern of changes in V observed in cycle 20 is not present in any of the other 13 cycles.
Statistical relationships between V, B and plasma density (N) with Ap are studied and it is shown that over three solar cycles (20, 21 and 22) the patterns are almost the same with a slight change observed in cycle 21. IMF B and Ap are linearly related over a wide range of Ap values from close to 0 up to about 60, whereas density/Ap relation appears insignificant. In case of V, an initial rapid rise in V causes moderate changes in Ap but for velocity in excess of 700 km/sec, the enhancement in geomagnetic activity is more rapid. From the statistical relation of several other solar wind/IMF parameters, their variability and combinations with Ap, one sees linear relation for solar wind electric field, n/s component of IMF and variability in the components of B. An estimate is made of the base level of the magnetosphere, corresponding to quiet levels of geomagnetic activity.

抄録全体を表示

A large M_w = 8.1 earthquake occurred off southeast coast of Antarctica near the Balleny Island region on March 25, 1998. We inverted teleseismic body-wave records to determine the rupture pattern using an iterative deconvolution method. The source parameters obtained are: the centroid depth=20km, (strike, dip, rake)= (282, 83, -1), the seismic moment M₀ = 1.6 × 10²¹ Nm (M_w = 8.1), the length L = 200 km, and the average slip D = 4.4 m. This earthquake occurred in the mid-plate but there has been no reports of such large earthquakes in this region. Furthermore, the source mechanism cannot be related to the plate motion inferred from the nearby transform faults. Therefore this earthquake is not a usual tectonic event. Here we show that the compressional axis of our source mechanism coincides with the horizontal crustal motion predicted by the Earth's response to present-day and past ice mass changes in Antarctica. Our result suggests that the 1998 Antarctica earthquake is caused by the postglacial rebound in the Antarctica.

抄録全体を表示

MF radar systems are able to determine horizontal neutral winds in the mesosphere and, to some extent in the lower thermosphere by cross-correlations of signals received at spaced antennas. Essentially, by also computing auto-correlations, signal fading may be measured which in turn is thought to be largely attributable to turbulence. Hitherto, estimates of upper limits for the turbulent energy dissipation rate have been derived from the characteristic fading times. In this paper, we propose that power spectra of the velocity components themselves may be used to yield estimates of turbulent energy dissipation rate. 2-minute resolution velocities from the Universities of Saskatchewan, Trom@s/os83/@ and Nagoya joint MF radar at 69°N, 19°E are used in a pilot analysis to illustrate and ratify the method.

抄録全体を表示