服部 克巳

A review of data processing of electromagnetic emission observation collected at the Complex Geophysical Observatory Karimshino (Kamchatka peninsula) during the first 5 months (July-November, 2000) of its operation is given. The main goal of this study addresses the detection of the phenomena associated with Kamchatka seismic activity. The following observations have been conducted at CGO: variations of ULF/ELF magnetic field, geoelectric potentials (telluric currents), and VLF signals from navigation radio transmitters. The methods of data processing of these observations are discussed. The examples of the first experimental results are presented. © 2001 European Geophysical Society.

Regular monitoring of some geophysical parameters in association with seismicity has been carried out since last year at the Japan-Russian Complex Geophysical Observatory in the Kamchatka region. This observatory was organized in connection with the ISTC project in Russia and was motivated by the results of the FRONTIER/RIKEN and FRONTIER/NASDA research projects in Japan. The main purpose of the observations is to investigate the electromagnetic and acoustic phenomena induced by the lithosphere processes (especially by seismic activity). The seismicity of the Kamchatka area is analyzed and a description of the observatory equipment is presented. At present, the activity of the observatory includes the seismic (frequency range δF =Combining double low line 0.5 - 40 Hz) and meteorological recordings, together with seismo-acoustic (δF =Combining double low line 30 - 1000 Hz) and electromagnetic observations: three-component magnetic ULF variations (δF =Combining double low line 0.003 - 30 Hz), three-component electric potential variations (δF ≤ 1.0 Hz), and VLF transmitter's signal perturbations (δF ∼ 10 - 40 kHz). © 2001 European Geophysical Society.

ULF emission data at Lunping (epicentral distance, 120 km) have been analysed for the Chi-Chi earthquake (with magnitude 7.6 and depth of 11 km) in Taiwan which occurred on 21 September 1999. Simple intensity analyses have not yielded any significant results but we have found, based on the analysis of polarization (the ratio of vertical magnetic field component Z to the horizontal component G), that the polarization (Z/G) showed a significant enhancement for two months before the earthquake. This kind of temporal evolution of polarization seems to be very similar to previous results, so that it is highly likely that this phenomenon may be associated with the Chi-Chi earthquake. Also, the comparison of the results of polarization analyses, by changing the signal threshold, has given us an approximate intensity of the seismogenic emission of the order of the monthly mean value. © 2001 European Geophysical Society.

Measurements of ULF electromagnetic disturbances were carried out in Japan before and during a seismic active period (1 February 2000 to 26 July 2000). A network consists of two groups of magnetic stations spaced apart at a distance of ≈140 km. Every group consists of three, 3-component high sensitive magnetic stations arranged in a triangle and spaced apart at a distance of 4-7 km. The results of the ULF magnetic field variation analysis in a frequency range of F = 0.002-0.5 Hz in connection with nearby earthquakes are presented. Traditional Z/G ratios (Z is the vertical component, G is the total horizontal component), magnetic gradient vectors and phase velocities of ULF waves propagating along the Earth's surface were constructed in several frequency bands. It was shown that variations of the R(F) = Z/G parameter have a different character in three frequency ranges: F1 = 0.1 ± 0.005, F2 = 0.01 ± 0.005 and F3 = 0.005±0.003 Hz. Ratio R(F 3)/R(F1) sharply increases 1-3 days before strong seismic shocks. Defined in a frequency range of F2 = 0.01 ± 0.005 Hz during nighttime intervals (00:00-06:00 LT), the amplitudes of Z and G component variations and the Z/G ratio started to increase ≈1.5 months before the period of the seismic activity. The ULF emissions of higher frequency ranges sharply increased just after the seismic activity start. The magnetic gradient vectors (B ≈ 1 - 5 pT/km), determined using horizontal component data (G ≈ 0.03 - 0.06 nT) of the magnetic stations of every group in the frequency range F = 0.05 ± 0.005 Hz, started to point to the future center of the seismic activity just before the seismoactive period; furthermore they continued following space displacements of the seismic activity center. The phase velocity vectors (V ≈ 20 km/s for F = 0.0067 Hz), determined using horizontal component data, were directed from the seismic activity center. Gradient vectors of the vertical component pointed to the closest seashore (known as the "sea shore" effect). The location of the seismic activity centers by two gradient vectors, constructed at every group of magnetic stations, gives an ≈ 10 km error in this experiment. © European Geophysical Society 2001.

Co-seismic signals of geoelectric potential have been observed for several earthquakes (EQs) in Japan. It has been argued that if there are precursory changes, stronger signals should be at the time of EQ. Our results prove that co-seismic signals do exist, but they start with the arrival of seismic waves and not at the origin time of EQs. Their amplitude does not scale with the dipole length. Both facts indicate that the changes are local effects of passing seismic waves and not due to electric signals emitted from the seismic source. Observed changes contain two components; oscillatory and offset/decay ones. The former is synchronized with seismic waves, whereas the latter, with a rise time of a few seconds and a decay time of 0.1 - 1 hour, is suspected to be related to EQ induced subsurface hydrological change. In both cases, the origin of electric potential variations may be electrokinetic.

ULF electromagnetic emissions associated with a large earthquake occurred at Biak Island, Indonesia at 5h 59m UT on February 17, 1996 (magnitude (Mw) = 8.2 and depth = 20 km from USGS catalog), have been investigated on the basis of ULF magnetic observations at two stations, Biak and Darwin in Australia (about 1,200 km apart). Though a simple analysis of magnetic field intensity (horizontal and vertical magnetic field components) at both stations was found to be closely correlated with geomagnetic ΣKp activity, a detailed analysis of the difference of H and Z components at the two stations, the polarization analysis (Z/H) and fractal analysis (frequency spectrum slope) at these two stations has yielded that the ULF emissions (in the frequency range from 5 mHz to 30 mHz) about 1.5-1.0 months before the quake are likely to be a precursory signature of the quake with its intensity on the order of 0.2-0.3 nT.

For the purpose of the direction finding of magnetospheric VLF/ELF radio waves, the reconstruction of the wave distribution function has been studied on the linear regularization of Phillips-Tikhonov type and on the associated optimization with the generalized cross-validation (GCV) in numerically analyzing spectral matrix data. On the ground-based wave observation, one finds by numerical simulation that the GCV well behaves in spite of the small number of data and that the linear algebra approach has the advantages of automatically selecting the best image and of computational facility. From the results of simulation the feasibility of direction finding is discussed. © 1995 Taylor & Francis Group, LLC.

Based on our previous experimental results of hiss-triggered chorus emissions and the theory of Helliwell, we propose a possible mechanism for how a chorus emission is generated from a wavelet in the hiss band. It consists of two steps : the phase bunching process of the resonant electrons and the radiation process. We select a VLF event observed onboard satellite GEOS1 on July 21,1977 and verify the proposed mechanism computationally. Whether the phase bunching occurs or not is examined to investigate the phase bunching time, the resonance time, and the duration of the wavelet. The validity of the generation model is considered through the reconstruction of chorus elements with use of three-dimensional ray tracing computations. Special attention is paid to the slope of the chorus elements and the frequency dependence of the wave normal angles for a chorus element. The theoretically reconstructed chorus element which satisfies the temporal variations onboard the satellite is found to occur with the Gendrin angle for the generation angle at the equator. Furthermore, there is a tendency for the generation angle to the Earth's magnetic field to shift to a larger angle with increase of frequency and to terminate around the oblique resonance angle. Propagation after the radiation is also found to be in the non-ducted mode through the computations.

磁気圏VLF/ELF波動の波動分布関数を用いた地上方位測定に,フィリップス-ティコノフ正則化法を応用した.従来,分布関数の再生には最大エントロピー法(MEM)が用いられてきたが,対数関数を目的関数とするため非線形な最適化計算を要し,計算時間が長く,再生された波動分布像の良さについて客観的な判定基準に弱点がある.そこで,本論文では目的関数として空間的な2次導関数を採用した線形計算法である上記の正則化法を応用することにより,計算の高速化を図ると共に,再生像の良さを評価するために,GCV最小化原理を導入した波動分布関数測定法を提案し,その有効性を1波源シミュレーションによって調査した.その結果,GCVが最小を示す場合に再生像は与えられた仮定像を最も良く復元しており,波源の中心位置も良い精度で推定していることが明らかとなった.つまり,最小GCV基準の導入により,客観的な最適再生像の選択が可能となった.また,線形正則化法の使用により,一つの最適像を得るための処理として,MEMと比較してほぼ1けたの計算高速化が実現した.

The results of detailed spectral analyses and direction finding for ELF data containing simultaneous hiss and chorus have been reported on the basis of the data observed onboard GEOS1 satellite in the outer magnetosphere at a high geomagnetic latitude (Λ～_25°). A combination of the present results with those from our previous equatorial studies, has yielded the following findings : (1) Each chorus element has a tendency to be originated from the hiss band and is asymptotic to the hiss band. (2) The intensity and occurrence of chorus are closely correlated with the intensity of the underlying hiss. (3) The hiss band exhibits some structures or wavelets (i. e. monochromatic wave components with significant duration), and we notice the existence of the causative wavelet at the foot of each chorus element. (4) There is a tendency that the intensity of wavelet increases with decreasing its duration. Furthermore, when we expect triggering chorus from a wavelet near the upper edge of the hiss band, the duration of the wavelet increases with decreasing its intenisty. (5) Very similar φ (azimuthal angle) values are noticed for both the hiss and chorus, suggesting that both phenomena come from the same source region. These observational facts might lead us to conclude that a wavelet existing near the upper edge of the hiss band is able to generate a chorus emission through coherent wave-particle interaction in the outer magnetosphere as in the case of active VLF wave injection experiments.

The results of detailed spectral analyses and direction finding for ELF data containing simultaneous hiss and chorus have been reported on the basis of the data observed on-board GEOS 1 satellite in the outer magnetosphere at a high geomagnetic latitude (Λ ∼- 25°). A combination of the present results with those from our previous equatorial studies has yielded the following findings : (1) each chorus element has a tendency to originate from the hiss band and is asymptotic to the hiss band ; (2) the intensity and occurrence of the chorus are closely correlated with the intensity of the underlying hiss ; (3) the hiss band exhibits structures or wavelets (i.e. monochromatic wave component with significant duration) and we notice the existence of the causative wavelet at the foot of each chorus element ; (4) when we expect triggering chorus from a wavelet near the upper edge of the hiss band, the duration of the wavelet increases with decreasing intensity ; (5) very similar φ (azimuthai angle) values are noticed for both the hiss and chorus, suggesting that both phenomena come from the same source region. These observational facts might lead us to conclude that a wavelet existing near the upper edge of the hiss band is able to generate a chorus emission through coherent wave-particle interaction in the outer magnetosphere, as in the case of active VLF wave injection experiments. © 1991.

The effect of the plasmapause on whistler-mode wave propagation and generation is investigated by means of the two-dimensional ray tracings started at the equator over a wide frequency range, 0.1 < Λo < 1.0 (Λ0: wave frequency normalized by the equatorial gyrofrequency) and for a wide range of the initial wave normal angle θ0. The main important findings are summarized as follows: (1) efficient gradient trapping is recognized at the inner edge of the plasmapause (especially with strong gradient) for lower frequency (Λ0 < 0.4) waves with smaller initial θ0s (-40° ∼ + 40°) generated at the equator; (2) lower frequency waves generated at the equator inside the plasmapause are found to have an access to low-altitude satellites and to ground, and their transmission latitude is slightly smaller than the plasmapause location; and (3) the outer edge of the plasmapause (especially with weak gradient) is also found to act as a guide for higher frequency (Λ > 0.5) waves. These theoretical findings are discussed with their relevance to the experimental facts of different magnetospheric whistler-mode emissions. © 1991.

The satellite and ground VLF data have indicated that chorus is often accompanied by a background hiss, and the purpose of this paper is to investigate the association between hiss and chorus and to present detailed experimental findings on the cause and effect relationship between hiss and chorus. The present study is based on detailed spectral analyses and direction finding measurements for VLF emission events with the simultaneous occurrence of hiss and chorus observed by GEOS-1 satellite in the outer magnetosphere. Two different events are analyzed : one is the equatorial (geomag. lat.=6.8°-8.5°) event, and the other the off-equatorial (24.2°-23.6°) event. The following important experimental findings have emerged from the present study : (1) Each chorus element originates from the hiss band and is asymptotic to the hiss band. (2) The intensity and occurrence of chorus emissions are closely correlated with the intensity of the underlying hiss band, such that when the intensity of the hiss band exceeds a threshold value of some mγ/√<Hz>, the excitation of chorus emission is enhanced. (3) Some parts of hiss band exhibit some structures or wavelets (i.e. monochromatic wave components with duration of more than 100ms), although hiss has been considered to be incoherent and turbulent. Furthermore, on some occasions, we can notice an existence of wavelets at the foot of chorus elements. (4) The direction finding results have newly indicated similar values for the azimuthal direction φ for both phenomena, suggesting that both hiss and chorus come from the same source region. These experimental facts suggest that a wavelet existing in the hiss band may generate a chorus emission through coherent wave-particle interaction in the outer magnetosphere as in the case of active VLF wave injection experiments.

The previous direction finding measurements of chorus emissions with different spectral shapes (rising, falling, and constant tones) near the equator have indicated that those chorus emissions are generated around the equator with their wave normals being nearly aligned with the magnetic field in terms of the Cyclotron instability by energetic electrons. Their subsequent propagation in the outer magnetosphere is discussed in the present paper. First, the results of wave normal directions of chorus observed in the off-equatorial regions are reviewed; one from OGO 5 measurement and the other, our GEOS 1 result. These measurements are compared with the results of two-dimensional ray-tracing analyses for a realistic model of the outer magnetosphere. Finally, it is found that the subsequent propagation of chorus after the equatorial generation, is, on some occasions, ducted, and it is non-ducted on other occasions.

The GEOS l satellite wave data have been used to determine the wave normal directions of chorus emissions in the off-equatorial region (geomagnetic latitude ∼ 17) of the outer magnetosphere based on the direction finding of the wave distribution function method. The local time and L shell of the data analysed are L.T. ≅ 10 h and L ∼ 7.60, respectively, and only one event is extensively studied. There are two frequency bands of emissions; one above and the other below one half the electron gyrofrequency, but the latter is the dominant emission and is investigated here. Two types of chorus are dealt with ; rising tones and impulsive ones. About 90% of chorus emissions analysed are found to be composed of a single plane wave, which is in contrast to plasmasphcric hiss. There is no conspicuous relationship between θ (the angle of wave normal direction with respect to the Earth's magnetic field) and frequency for rising tones, and their θs take larger values in a range of 30°-55, whereas a tendency seems to appear, for an impulsive charus. for waves at higher frequencies to travel at a larger angle to the field. The wave normal directions observed have been discussed as functions of wave frequency normalized by the local electron gyrofrequency and the slope of an emission, df{hook} dt. A comparison of the present results on the off-equatorial wave normal directions with the previous equatorial direction finding results have made it possible to discuss the generation and propagation mechanism of chorus emissions with different structures including rising tones, falling tones, constant frequency tones and impulsive (burstlike) ones, and finally we have pointed out future work for investigation. © 1990.

The co-existence of hiss and chorus is frequently observed on board satellites in the outer magnetosphere, and chorus emissions are observed in the data from GEOS-1 satellite to start nearly at the uppermost frequency of the hiss band. The direction finding measurements have been made for these hiss-triggered chorus events. On the basis of the direction finding results and detailed analysis of the fine structures in the hiss band, we have suggested a generation model how a chorus is triggered by a hiss such that some coherent wavelets near the top of the hiss band phase-bunch electrons at the equator, and these phasebunched electrons radiate a coherent signal as they move away from the equator. Also, theoretical computations have been carried out in order to explain the df/dt, θ value (the polar angle between the wave normal direction and the Earth's magnetic field) and other factors by means of a two-dimensional ray-tracing method. The differences between the experimental and theoretical values are pointed out, and some possibilities to remedy these discrepancies are discussed.

In order to have a clue to the study of a link between the hiss and chorus, including the mechanism of the hiss-triggered chorus, statistical properties of magnetospheric VLF/ELF hiss including the stationarity, ergodicity, normality and monochromaticity (coherence or incoherence), have been investigated on the basis of the VLF/ELF hiss data observed at Moshiri in Japan and at Brorfelde in Denmark. It is then found that magnetospheric VLF/ELF hiss is generally considered as a Gaussian, stationary random signal. However, a difference from the white noise generated by a random noise generator is found such that there are some kinds of structures such as coherent (monochromatic) large amplitude spectral components (wavelets) existing within the hiss band.

The GEOS-1 satellite wave data have been used to determine the wave normal directions of chorus emissions in the off-equatorial region of the outer magnetosphere based on the wave distribution function method. The wave normal directions with respect to the magnetic field have been discussed as functions of wave frequency normalized by the local gyrofrequency and of the slope, df/dt of an emission. Two different kinds of chorus structures (rising tone and impulsive one) have been treated. A comparison of the present behaviour of off-equatorial wave normal directions with the previous equatorial wave normal measurements has enabled us to discuss the generation and propagation mechanism of chorus emissions with different structures (rising tone, falling tone, constant frequency tone and impulsive one), and finally we suggest the future problems to study.