J.Geomag. Geoelectr.,42,459-478,1990
VLF Observations by the Akebono(EXOS-D)Satellite
Iwane KIMURA1,9, Kozo HASHIMOTO2, Isamu NAGANO3, Toshimi OKADA4,Masayuki YAMAMOTO1, Takeo YosHINO5, Hiroshi MATSUMOTO6, Masaki EJIRl7,
and Kanji HAYASHI8
1Department of Electrical Engineering II, Kyoto University, Kyoto 606, Japan2Department of Electrical Engineering, Tokyo Denki University, Tokyo 101, Japan3Department of Electrical Engineering, Kanazawa University, Kanazawa 920, Japan4Research Institute of Atmospherics, Nagoya University, Toyokawa 442, Japan
5Department of Communication, Electro-Communication university, Tokyo 182, Japan6Raclio Atmospheric Science Center, Kyoto University, Uji 611, Japan
7National Institute of Polar Research, Tokyo 173, Japan
8Geophysics Research Laboratory, University of Tokyo, Tokyo 113, Japan9Institute of Space and Aastronautical Science, Sagamihara 229, Japan
(Received August 1,1989;Revised December 15,1989)
The VLF wave instruments on board Akebono(EXOS-D)involve a wide bandanalyzer(WBA), multi-channel analyzers(MCA), Poynting flux analyzers(PFX), ELFfrequency analyzers and a vector impedance probe(VIP)and cover a frequency rangefrom a few Hz to 17.8kHz for electric and magnetic field components. The mainobjectives of the wave observations are to investigate the wave phenomena closelyassociated with energetic particle precipitation in the auroral zone and the physics ofwave-wave and wave-particle interaction phenomena. In order to clarify these wavecharacteristics, the determination of the wave normal direction and the direction of thePoynting vectors are essential. Our VLF instruments, especially PFX and ELF, aredesigned to satisfy such requirements. The present paper introduces all the subsystems ofthe VLF instruments and some preliminary results of observations to show the charac-teristic features of each subsystem.
1. Introduction
The EXOS-D satellite was launched at 23:30 UT on February 21,1989 in order toinvestigate energy flow from the magnetospheric tail to the auroral region, and is now
named Akebono, which means dawn in Japanese. The VLF instruments on board
Akebono are designed to investigate the behavior of plasma waves associated with
accelerated auroral particles, wave particle interaction mechanisms and propagation
characteristics of whistler mode, ion cyclotron mode, and electrostatic mode waves in the
magnetospheric plasma.
Specific wave phenomena of interest in the auroral region are: i)AKR, ii)auroralhiss, iii)saucers, and iv)broad-band electrostatic noise. Item i)is a free space mode
phenomenon, ii)and iii)are whistler mode phenomena and iv)is an electrostatic mode
phenomenon. These phenomena are qualitatively understood. However, the sourceregion for each phenomenon and the generation mechanisms have not yet been clearly
understood.
Auroral hiss is generated in a loes-altitude auroral region and electrons in the keV
energy range appear to be responsible for the emissions. AKR and auroral hiss appear to
459
460 I.KIMURA et al.
be generated on the same field line(GURNETT et al.,1983;CALVERT and HASHIMOTO,
1989).
Saucer observations by the ISIS satellites suggest that the phenomenon is generated
in the 500 to 3500km altitude region. It is, however, unclear why the saucer is generated
in such a region. Absorption or intensified bands at the harmonics of the ion cyclotron
frequency are also observed in saucer spectra(HORITA and JAMES,1982). This
phenomenon is also not yet completely understood.
The broad band electrostatic noise observed by the Hawkeye 1 and Imp 6 satellites
suggests that this noise occurs on the same magnetic field lines as the inverted V electron
events(GURNETT and FRANK,1977). Some propagation characteristics of the noise
observed by the S3-3 satellite were investigated by TEMERIN(1978).
Another electrostatic mode, electrostatic hydrogen cyclotron waves found by S3-3
near 1 RE(MOZER et al.,1977), has been shown to be consistent with the electrostatic ion
cyclotron dispersion relation(KINTNER et al.,1978).
As for the so-called wave-particle interaction phenomenon, represented by ASE
(artificially stimulated emissions), it has been found by the Jikiken observations of Siple
station transmissions that ASE's were generated by intense fluxes of resonant electrons at
small pitch angles(KIMURA et al.,1983). However, the wave intensities associated with
the ASE phenomenon observed by several satellites seem to be much lower than those
theoretically required for the wave particle interaction(INAN et al.,1977). ISIS satellite
observations of Aldra Omega signals and emissions triggered by these signals revealed
that the wave normal angle at the magnetic equatorial plane was 10-30°, but was
sometimes much larger(MATSUO et al.,1985).
Information on the wave normal directions of the triggering wave for triggered
emission events is very important for the study of the mechanism of wave-particle
interactions.
For a study of the propagation characteristics of whistler mode signals transmitted
from the ground-based VLF transmitters, such as Omega signals, the delay time and the
direction of the wave normal vectors of the signals are measured on board Akebono.
Electron density at the satellite orbit, which is needed for theoretical confirmation of the
characteristics, e.g. by the ray tracing technique(KIMURA et al.,1985;HASHIMOTO et al.,
1987),is determined by PWS(high frequency plasma wave analyzer).
In the present paper, first the subsystems of the VLF instruments are briefly
described and then some preliminary results of observations are introduced.
2. Description of the Subsystems of the VLF Instruments
In order to satisfy the above mentioned requirements, instruments are needed to
measure the wave normal and Poynting flux directions of these wave phenomena in
addition to their dynamic spectra. Vector electric as well as magnetic field components in
the frequency range from a few Hz to MHz must also be measured. The frequency range
above tens of kHz is covered by the PWS, the other wave instrument of Akebono for
high frequency plasma wave phenomena.
The VLF is responsible for the frequency range below 17.8kHz down to 3.16Hz,
and is composed of loop and dipole antennas, common preamplifiers directly connected
with the sensors, and the following subsystems: WBA(wide band analyzer), MCA(multi-
channel analyzers), PFX(Poynting flux analyzers), ELF(ELF range analyzers)and VIP
VLF Observations by the Akebono (EXOS-D) Satellite 461
Fig. 1. Block diagram of VLF instrument.
Fig. 2. Configuration of wave sensors and coordinate systems.
462 I.KIMURA et al.
(vector impedance probe). The block diagram of the VLF instruments is shown in Fig.1.
In this figure, the so-called DPU common is comprised of buffer amplifiers between the
sensor sections and the above-mentioned subsystems.
2.1 Sensors and their coordinate systems
To detect the electric component of the plasma wave fields, a pair of crossed dipole
wire antennas,60m tip-to-tip shielded for 3m nearest to the spacecraft, are used. For the
magnetic field sensors, three orthogonal open loop antennas installed on the tip of a 1.5
m mast are used for frequencies higher than 800Hz. For frequencies lower than this
frequency limit, three orthogonal search coils are used, which are mounted on the tip of
the search-coil mast of 3m length.
The coordinate systems for the various antennas and search coils are shown in Fig.
2. First of all the spin axis of the satellite, which is always directed to the sun with an
accuracy within 2.5°, is taken to be the Z axis. The mast of the loop antennas is directed
in the positive Y direction and the two masts of 3m and 5m for magnetic flux gate
magnetometers and search coils are directed in the -X and -Y directions respectively as
shown in Fig.2. The three orthogonal flux gate magnetometers and search coils measure
the X, Y, and Z components of the earth's magnetic field and their induction in the above
satellite coordinate system, and the geomagnetic aspect sensors(GA)also refer to this
coordinate system. The outputs of the three search coils are named sBx, sBY, and sBz.
The plane normal directions for the three loop antennas are orthogonal to each
other, but are directed in different directions from the X, Y, and Z axes so that the output
of the loop antennas are called B1, B2, and B3. The relation between the magnetic
components(B1, B2, B3)and those components in the satellite coordinates,(BX, BY, BZ)is
given by
The directions of the crossed dipole wire antennas are directed about 45° off from
the X and Y directions and the two components of the electric fields are called Ex and Ey
where the positive x and positive y axes are in the first and second quadrants in the X-Y
plane with the x axis 35° off from the X axis.
The equivalent noise level of the preamplifier for electric field measurement is about
4.5×10-8V/√Hz in the major part of the VLF band when the input terminal of the
preamplifier is grounded to the chassis of the satellite.
As to the sensitivity of the magnetic sensors, the open loop antennas are rectangular
(60cm×60cm)coils with 10 turns each, and the frequency dependence of the effective
length of each loop including stepup by a transformer and a preamplifier is shown in Fig.
3.These loop antennas were designed to be used for VLF as well as PWS(for the higher
frequency)by using two independent transformers whose primary coils are connected in
VLF Observations by the Akebono (EXOS-D) Satellite 463
Fig.3. Frequency dependence of the effective length of the loop antennas and search coils.
series with the loop antenna circuit and whose secondary coils are independently
connected to the two preamplifiers for VLF and for PWS(OKADA et al.,1987). The
magnetic field detection threshold level at 5kHz is 0.005pT/√Hz.
The search coils are composed of 105 turns of 50μm polyurethane wire on a super-
permalloy rod of 3×3×300(mm); the sensitivity of the coils is 0.2pT/√Hz at 100Hz.
The frequency dependence of the effective length is also shown in Fig. 3.
2.2 WBA(wide band receiver for observation of VLF spectra)
Wide band analog VLF signals received in the frequency range from 50Hz to 14
kHz or 50Hz to 7kHz(selectable by command)are transmitted directly to the ground by
analog telemetry. The field components to be measured(Ex, Ey, sBY or B2)are selectable
by command. The gain of the WBA receiver is changed in 25 dB steps from 0 dB up to 75
dB by checking the averaged signal level every 0.5s. This procedure is automatically
executed by the WIDA hybrid IC, specially designed for the VLF instruments of this
satellite, by which more than 80 dB of dynamic range is secured. Signal intensity can also
be determined within the above dynamic range as long as the amplifier does not saturate.
2.3 MCA(multi-channel analyzers)
Sixteen MCA channels for Ex or Ey(selectable by command)and 16 channels for
sBY and B2(chs.1-10 for sBY and chs.11-16 for B2)are included in the VLF electronics.
The center frequency for each channel is 3.16,5.62,10.0 and 17.8[Hz]×1, ×10, ×100 and
×1,000 with a band width of 30% of each center frequency. Relative filter response curves
for 16 channels are shown in Fig.4, which is very similar to the ISEE multi-channel
analyzer(SCARF et al.,1978). The maximum allowable input voltage is 1 Vrms and a
dynamic range of more than 80 dB is obtained by the WIDA IC. The detector output is
converted to an 8 bit word and PCM-transmitted with sampling rates of 2 per s for the
high(H)bit rate,4 per s for the medium(M)bit rate, and 1 per s for the low(L)bit rate
transmission.
464 I. KIMURA et al.
Fig.4. Frequency response characteristics of MCA channels.
2.4 PFX(measurement of wave normal direction and Poyntingflux)Five components of electric and magnetic fields, Ex, Ey,(B1 or sBX),(B2 or sBY), and
(B3 or sBZ)are 12-bit-A/D-converted and log-compressed to 8 bits at a sampling rate of320Hz and sent to the ground by the PCM telemetry. The wave normal direction and thePoynting flux of received signals are calculated on the ground. The PFX is composed offive channels of triple-super-heterodyne receivers with an output band-width of 50Hz, asshown by a block diagram in Fig.5. The local oscillator frequency is either automaticallystepped by the minimum step of 50Hz or kept fixed at a constant frequency in thefrequency range from 100Hz to 12.75kHz. The selection of stepped or fixed mode ismade by command while the selection of B or sB is made automatically according to theselected center frequency. The dynamic range of the PFX is also more than 80dB, byusing the WIDA IC, which is illustrated by WIDA's in the block diagram shown inFig.5.
2.5 ELF(ELF range receiver)The ELF range receiver can be operated in either the 4 channel or 2 channel mode,
selectable by command. In the 4 channel mode, waves in a frequency range less than 50Hz are observed and are sampled by 160 Hz to 8 bit words. The four channels arecomposed of ch.1; Ex or Ey(selectable by command), ch. 2; sBX, ch.3; sBY, and ch.4;sBZ). In the 2 channel mode, the upper limit frequency is increased up to 100Hz with asampling frequency of 320Hz with 8 bit resolution, but only two components selected bycommand can be observed. The dynamic range for either case is 80dB. The wave normal
direction and the Poynting flux is then calculated by the ground data processing.
2.6 VIP(measurements of vector impedance of the wire antennas)In order to determine the electric field intensity of a signal as accurately as possible,
the antenna impedance must be known. Figure 6 shows the principle of the measurement.Aconstant current source is applied to the input terminals of the balanced preamplifier
VLF Observations by the Akebono (EXOS-D) Satellite 465
466 I. KIMURA et al.
Fig.6. Block diagram and equivalent circuit of VIP.
(Fig.6(a))with the center of the preamplifier being grounded to the chassis of thesatellite, and the relative output amplitude and phase with respect to the applied signalare measured with the VIP electronics to determine the vector impedance of the antenna.Figure 6(b)illustrates one side of the antenna and preamplifier system relative to thechassis of the satellite. Then RS and Cs+Cin, where RS and CS are the sheath resistance andcapacitance of the antenna respectively, and Cin is the input capacitance of the pre-amplifier relative to the ground, can be measured. Cin measured before launch was about100pF. The applied source intensity is controlled by feed-back so as to keep the outputvoltage constant. This makes it possible to measure the impedance in a wide range.Figure 6(c)is an equivalent circuit for signal reception. E is the electric field intensity ofthe signal and heir is the effective length of the antenna. The input signal to the pre-amplifier is divided by the sheath impedance and the input capacitance as shown in Fig.6(c).One of the antenna pairs, that is Ex or Ey, and the frequency for the impedancemeasurement is selected by command.
2.7 DPU(data processing unit)The main function of DPU is to provide an interface between the spacecraft data
handling unit(DHU)and the VLF instruments. The DPU also provides the controlsignals for the subsystems of the VLF instrument.
VLF Observations by the Akebono(EXOS-D)Satellite 467
3. Results of Initial Observations
Our VLF observations started when the wire antennas, loop and search coils were
deployed in the first week of March, about 10 days after launch, while most otherscientific instruments using high voltages were turned on at the end of March. WBA data
via U band telemetry have been taken at Kagoshima Space Center(KSC)and Prince
Albert(PA)in Canada. Other VLF data via PCM telemetry have been taken at KSC,
PA, Syowa in Antarctica, and Esrange in Sweden.
In the following, some of the analyzed results of each VLF subsystem are introduced
to show the characteristics of all subsystems described in the previous section.
3.1 Wide band spectra of various VLF phenomena including emissions triggered byOmega signals
Wide band VLF spectra taken by WBA are found to be of high quality with a
sufficient S/N ratio realized by the WIDA IC.
A couple ofexamples of spectra taken by WBA are shown in Panels 1 and 2. Panel 1
illustrates an example of f-t spectra including whistlers, LHR noise, ground-based VLF
signals at 10.2,11.05 and 11.9kHz, observed by one pair of the 60m tip-to-tip wire
antennas. The location of the satellite was at an altitude of 2800km, a geomagnetic
latitude and longitude of 40.0°N and 163.5°W respectively, and an L value of 2.7.
Panel 1. An example of wide band VLF spectrum of whistlers, LHR noise, and ground based VLFtransmitter signals.
468 I.KIMURA et al.
Signal strength of wave phenomena at any particular frequency and time can be read
on the f-t display on the CRT after FFT processing. In the case shown in panel 1, the
maximum signal intensity at the input terminal of the preamplifier was 7.85mV for the
most intensified portion of LHR noise with a whistler at 6.1kHz and at 10s after the
start of the spectrogram. lf we assume that the effective length of the antenna is 27 m and
the pickup factor is 0.75, the electric field is 0.39mV/m. As seen on the panel, the signal
intensity is strongly modulated by the satellite spin motion with a period of 8s, because
of the electrostatic characteristics of the LHR noise. Fortunately, the spin axis is directed
to the sun to make the solar cells always face the sun, so that there is no shadowing effect
to cause additional electrostatic noise at multiple harmonics of the spin frequency.
Panel 2 illustrates one example of spectrograms of observed Omega signals
transmitted from Australia(geomagnetic latitude of 47°S), which shows a sequence of
Australian Omega format; 11.05,13.0,10.2,13.6,11.333,13.0kHz. This spectrogram
was obtained by the loop antenna and VLF-WBA on April 5,1989. As seen on the panel,
the Omega pulses at frequencies of 10.2,11.05 and 11.333 are accompanied by strong
triggered emissions(so-called ASE). On this pass, the Omega signals were observed in the
period from 07:30 UT to the end of telemetry reception, i.e. 07:40 UT, while from 07:35
UT until 07:40 UT, strong emissions (ASE's)associated with Omega signals were
observed.
The delay of the leading edge of the received Omega pulse at 10.2kHz, from that of
the signal transmitted can be measured on the spectra with an accuracy of 20ms. A delay
time of the telemetry signal to the tracking station, which is about 20ms for a range of
6,000km, must be compensated by subtraction to obtain the virtual delay time of Omega
signals. The delay times(not corrected)for the above-mentioned period are in the range
from 0.5 to 0.7s, as shown in Fig.7. It was found by ray tracing that these delay times are
reasonable values for a direct single path to the satellite from the Omega transmitter in
Australia.
Fig. 7. The location of Akebono when Australian Omega signals and their triggered emissions were observed.
VLF Observations by the Akebono (EXOS-D) Satellite 469
Panel 2. VLF wide band spectrum of ASE's triggered by Australian Omega signals.
Panel 3. One example off-t spectrogram obtained by MCA, showing a funnel type auroral hiss emission in the auroral region (PCM data acquisition made at Prince Albert).
470 I.KIMURA et al.
The location of Akebono corresponding to the above time interval is illustrated in
Fig.8. After O7:35 UT for every pulse at 10.2kHz, the Omega signals are followed by
strong triggered emissions, with a delay of about 450ms from the leading edge of the
pulse. It is very interesting to note that the signal intensity at the transmitter frequency of10.2kHz suddenly increases after the delay of 450ms from the leading edge of each pulse.
Fig.8. Spatial variation of the delay time and the signal intensity of Australian Omega signals at 10.2kHzobserved on April 5, 1989.
In order to compare the signal intensity before the time when the emissions aretriggered with that after the triggered emissions, we have measured the peak signalintensity over 200ms just after the leading edge of the pulse(denoted by Ai)and the peakintensity over the interval on the pulse after the delay of 450ms(denoted by Am)while theAustralian Omega signals were detected, i.e. from 07:30 to 07:40 UT, including thenontriggering period(from 07:30 to 07:35 UT).
Figure 7 shows a time variation of Ai and Am, where the ordinate indicates therelative signal intensity, but the 0 dB on this figure corresponds to 1.1 pT in the magneticfield intensity. In the figure, Ai shows a tendency to slowly increase with time orincreasing latitude. This tendency may not be due to amplification, but may be mainlydue to a spatial variation of signal strength by a difference in propagation path from thesource to the satellite location. After 07:35 UT, Am, the signal strength after thetriggering, remarkably increases compared with Ai, implying that an amplification effectis functioning due to wave-particle interaction. The difference between Aiand Am is a
VLF Observations by the Akebono(EXOS-D)Satellite 471
measure of total amplification, which is about 24 dB at 07:40 UT. Actually, the telemetry
reception was terminated at the maximum triggering activity, so that the total amplifica-
tion could be larger than 24 dB, which is consistent with the results of ground
observations for the Siple transmitter(HELLIWELL and KATSUFRAKIS,1974).
on board Akebono we have already detected five other triggering events for the
satellite location in the northern hemisphere, in the interval of 09 to 10 UT on March 10,
11,21,and 27 for Australian Omega signals, and around 12:45 UT on March 27 for a
ground VLF signal at a frequency of 11.9kHz, which is not a frequency of any of the
Omega transmitters. These six events are concentrated within a month after March 10, so
that these triggering activities might be attributed to a period of high solar activity in
March and also to the local time of the satellite trajectory around evening or night, which
was a favorable condition for the signal to penetrate through the bottom of the
ionosphere at the source region.
3.2 MCA spectrogram
Each of the data from the 16 channels of MCA for E and B field components yields a
peak value every 1s and an average value every 0.5s. The MCA data are able to provide
us with general characteristics of the VLF wave activity, although the frequency
resolution is not good. In Panel 3, colored MCA spectrograms for E and B components
are shown for the data taken on March 11,1989 at Prince Albert, Canada, one of the
tracking stations for Akebono. The spectrogram is produced by drawing a contour map
based on the maximum of the averaged values for 4-s duration in time at 16frequencies.
Strong emissions, which are red-colored on the panel, covering all the observed
frequency range for the E and only down to 2kHz for the B component around 06:39 UT
correspond to a typical example of funnel-type auroral hiss emissions(GURNETT et al.,
1983).As shown in the panel, the emission was observed at an invariant latitude of 69°,
an altitude of 2,700km, and a magnetic local time of 20h.
3.3 Determination of wave normal directions
The PFX instrument can measure the five components consisting of three magnetic
fields and two electric fields in order to obtain the wave normal direction and Poynting
vector finally. It is not necessary to measure all six components to calculate the wave
characteristics for the electromagnetic waves because any one component can be
obtained from the measurement of the other five components, if the local electron density
is known(SHAWHAN,1970).
In the following, we will show one example of the determination of the wave normal
direction for whistlers, by analyzing the PCM data of the three entire magnetic
components. The wave normal direction can be calculated by a simple method based on
the facts that the wave normal direction of the electromagnetic waves is always
perpendicular to the polarization plane of their magnetic field, and that the observed
wave packet is composed of a plane wave with a single wave normal direction.
Figure 9(a)is an f-t diagram observed by WBA using one of the loop antennas. A
couple of whistlers shown were observed in the northern hemisphere at 10:53:18 UT on
March 24, 1989, at an altitude of 4396km and at a geomagnetic latitude of 21.8°,the first
one with a dispersion of 40s1/2 and the second about 60s1/2. Figure 9(b)indicates wave
forms of the five components(2 components of E fields and 3 components of B fields in
their coordinate systems of the antennas)of the whistlers at 6kHz with a bandwidth of 50
472 I. KIMURA et al.
Panel 4. ELF emissions observed in the magnetic equatorial region. The red lines represent the helium
cyclotron frequency at the satellite orbit.
VLF Observations by the Akebono (EXOS-D) Satellite 473
Fig.9.(a)f-t diagram of whistlers to be analyzed.(b)Wave form of the whistlers.(c)Direction of the wave
normal of a whistler.
Hz. The full length of the horizontal line corresponds to 0.5s. The wave form
corresponding to a whistler lasts about 30 ms. We can recognize two whistlers within the
above 0.5s. Figure 9(c)shows the wave normal direction with reference to the direction
of the earth's magnetic field which was simultaneously measured by the MGF instrument
on board the satellite. The calculated wave normal angle is nearly constant with an angle
of 60° during the period of strong signal strength, which is about 30ms. For the second
whistler in Fig.9(b)the wave normal angle was about 75~90°. The azimuth direction of
the wave normal vector can also be calculated if the satellite attitude is definitely
determined. Since the present paper is intended to show the preliminary results, only the
wave normal angle with reference to the earth's magnetic field is shown.
The source of whistlers with a dispersion larger than 20s1/2 is considered to be
located in the southern hemisphere. The large wave normal angle suggests that the wave
474 I.KIMURA et al.
has propagated to the satellite from the southern hemisphere in the non-ducted mode.
The above second whistler with a slightly larger dispersion than the first whistler is
interpreted in such a way that after passing through the satellite the first whistler
propagated down and was observed again after reflection either at the bottom of the
ionosphere or LHR reflection somewhere below the satellite.
We can get other characteristics of the wave from the five field components. In the
above whistler events, the polarization was of course right-handed circularly polarized
and the intensity of the magnetic component was 0.6pT for the band width of 50Hz. On
the other hand the electric field intensity was 10.2mV/m for the same band width, under
the assumptions that the effective length of the wire antenna is 27m, half of the tip-to-tip
length, and that the pickup factor calculated by using the sheath capacitance of the
antenna(see Subsection 3.5)is 0.75. The value of c|B|/|E| calculated by the above
quantities is l5.6, which should be the refractive index, ifthe |E| is the magnitude of
electric field perpendicular to both the wave magnetic field and the wave normal
direction. On the other hand, according to the on board observation by PWS and MGF,
the plasma and the cyclotron frequencies are 566kHz and 200kHz respectively. The
refractive index based on these parameters becomes 23 for the above mentioned wave
normal angle of 60°.The refractive index deduced from the field ratio and that obtained
from calculation is not exactly the same, but the difference is within a reasonable range if
we take account ofthe facts that we do not observe the Z component ofthe electric field
and the assumption of the effective length of the antenna is not always definitely
acceptable(SONWALKAR and INAN,1986).
3.4 ELF emissions observed near the magnetic equatorial plane
In the ELF range, which is below a frequency of 100Hz, there are many interesting
wave phenomena associated with ion constituents. One such phenomenon is the ELF
emissions that are observed only in the near vicinity of the magnetic equatorial plane.
Panel 4 shows one example observed on May 27,1989, when Akebono was approaching
the geomagnetic equator from the southern hemisphere. On the panel, four f-t spectro-
grams are shown for Ex and sBX, sBY, and sBZ components respectively. Strong
horizontal lines at frequencies of 32 and 64Hz are due to interference caused by other
instruments on board the satellite.
On each panel, the cyclotron frequency of helium at the satellite location is plotted
with a red line for reference. It is clearly seen that the frequency of emissions is always
above the helium cyclotron frequency(fHe)and is much below the proton cyclotron
frequency (fH). Moreover, there is a sharp frequency gap between the bottom of the
emissions and fHe. From this point, with the fact that this phenomenon has a magnetic
field component as well as an electric field component, it is conjectured that this emission
may be identified as one of the electromagnetic ion cyclotron modes existing above the
lower hybrid frequency by helium ions as indicated by the class 3 mode(according to the
classification by RAUCH and ROUX,1982)in the ω-k diagram as shown in Fig.10. This
mode is known to be confined to a narrow region around the magnetic equatorial plane
by LHR reflection(RAUCH and ROUX,1982), which was observed by GEOS 1 and 2
(YOUNG et al., 1981; ROUX et al., 1982).
In July,1989 several spectra were newly found in which two frequency bands coexist
between the local proton and helium ion cyclotron frequencies and between the local
helium and oxygen ion cyclotron frequencies. This phenomenon can be interpreted in a
VLF Observations by the Akebono (EXOS-D) Satellite 475
Fig.10. ω-k diagram of the electromagnetic ion cyclotron modes. The abscissa and the ordinate are
normalized by the proton cyclotron angular frequency(ΩH).
similar way as those having one frequency band above the helium cyclotron frequency, as
mentioned previously. If we can assume that there is a sufficient content of oxygen ions,
say a few percent, in addition to proton and helium ions, one more LHR resonance
frequency appears above the oxygen cyclotron frequency and below the helium cyclotron
frequency in the ω-k diagram in the presence of the three ion constituents.
Our calculation has shown that the presence of high temperature ions can amplify
these modes, so that the appearance of such ELF emissions in the magnetic equatorial
region may be an indication of the existence of high temperature ions there. Unfortunately,
only a limited number of measurements of energetic ion spectra could have been made in
the low latitude region, because the radiation belt particles cause high erroneous counts
to the detectors. Much more statistical study of this phenomenon and the measurements
of energetic ions are needed.
3.5 Vector impedance of the wire antennas in the VLF range
So far several VIP measurements have been analyzed. Typically CS=250 pF and
RS=500kΩ at altitudes higher than four thousand km. The antenna becomes resistive
below the frequency f0=1/(2πCSRS)=1300Hz, and capacitive at frequencies above f0.
Near this frequency, effects of both RS and CS must be taken into consideration. Observed
impedances did not vary so much. The capacity of the co-axial capacitor formed by the
antenna and the boundary of the plasma sheath is(AGGSON and KAPETANAKOS,1966)
(2)
where λ is the sheath thickness, a is the antenna radius(0.19mm), and d is the antenna
length(30m). C is about 270pF if the sheath thickness is assumed to be 10cm(typical
476 I.KIMURA et al.
Debye length in these regions). Therefore the above results are reasonable.
4. Discussion and Conclusions
We have so far confirmed that all subsystems of the VLF instruments on boardAkebono are working satisfactorily. The three components of the magnetic field and thetwo components of the electric field of electromagnetic plasma waves in the magnetospherecan be measured by the PFX subsystem using the crossed loop antennas or search coils atafixed frequency or by staircase sweeping the frequency. WBA uses one of the wiredipole antennas, or one of the crossed loop or search coil antennas.
In any subsystem the magnetic field intensity can be measured very accurately,because there is no ambiguity in the pickup factor of the sensors. For the determinationof electric field intensity, the E field channels provide us with signals with sufficientlyhigh S/N ratio, and we can also measure the sheath impedance of wire antennas asmentioned in Subsection 3.5. We therefore determine the electric field intensity indepen-dently from the magnetic field component. However, we know that there is an ambiguityin determining the electric field, such as an ambiguity of the effective length of the wireantennas(SONWALKAR and INAN,1986). We have, therefore, to collect enough paireddata of E and B fields, before we can use the electric field intensity with much confidence.
Aphase difference between any two channels for the measurement of E and B fieldsis also very important data necessary to be compensated in the calculation of the wavenormal direction. It was found that the phase difference between any two B channels isnegligible within a few degrees. The phase differences between E and B channels are notalways negligible but we are able to calibrate them by on board calibration.
It will take time for us to reach the situation where we can calculate the direction ofthe wave normal and the Poynting vector by using five components of the E and B fields.In the present paper, we have, therefore, introduced only one result of the wave normaldirection for whistlers based on the simple method, as mentioned in Subsection 3.3.
For the study of wave particle interaction phenomena, we have detected severalOmega triggering events, some of which are analyzed in the present paper. We still havemuch work to do to know the characteristics of the triggering effect. Moreover it is, ofcourse, necessary for us to observe energetic particles at the same time. However, so farwe have had not much chance for simultaneous observations in the middle and low
geomagnetic latitudes,. because in these regions high energetic particles in the radiationbelt cause tremendous error in measuring low energetic particle fluxes and deteriorate thechannel electron multiplier quickly. But such joint observations will be made in the nearfuture for our study.
In the ELF range, we have found very interesting ELF emissions which are observedin a limited range around the geomagnetic equatorial plane with frequencies above thehelium ion cyclotron frequency and sometimes above the oxygen cyclotron frequency aswell. These phenomena appear to be of the electromagnetic ion cyclotron modes existingabove the lower hybrid frequencies by helium ions and oxygen ions and trapped in thelower latitude zone by LHR reflection. So far there have not been many observations of
energetic electrons and ions in the geomagnetic equatorial region, unfortunately. Wecould not have confirmed the mechanism of generation mentioned in the present paper.The results of the detailed analyses of these phenomena will be published elsewhere in thenear future.
VLF Observations by the Akebono(EXOS-D)Satellite 477
We would like to thank Prof. K. Tsuruda, the Project manager, Profs. H. Oya and A. Nishida,Project Scientists, and the ISAS engineering team headed by Drs. M. Natori, I. Nakatani and J.Onoda at the project engineers, especially Mr. T. Oshima for their great effort in assembling andlaunching the satellite Akebono. We also have to acknowledge the staff of the overseas trackingstations at Syowa in Antarctica, Prince Albert in Canada and Esrange in Sweden for continualdata acquisition of Akebono telemetry.
We are also greatly indebted to NEC for their perfect arrangement of all satellite systems andtheir fabrication of the common part of the satellite, to Meisei Electric Co. Ltd. for theirfabrication of all subsystems of VLF, and to Nihon Hikoki Co. Ltd. for their fabrication of themasts for the search coils. We are also grateful to all colleagues of the Akebono scientificinstrument team for their cooperation and great effort in minimizing RF interference to our VLFinstruments.
In our data analyses, the electron density measured by PWS group and the magnetic fieldmeasured by MGF group were indispensable. We are grateful to these groups for their data used inour analyses.
Finally we acknowledge the cooperation and contribution of graduate students of Kimura'slaboratory and Nagano's laboratory in the preparation of software for satellite data handling andin the data analyses, especially Messrs. Akira Sawada, Yoshihiko Ito, Yoji Kishi and YoshiyaKasahara of Kyoto University and Eiichi Kennai of Kanazawa University.
REFERENCES
AGGSON, T. L. and C. A. KAPETANAKOS, On the impedance of a satellite borne VLF electric field antenna, X-612-66-380, NASA Goddard Space Flight Center, August 1966.
CALVERT, W. and K. HASHIMOTO, The propagation modes and propagation properties of auroral radio emissions, submitted to J. Geophys. Res., 1989.
GURNETT, D. A. and L. A. FRANK, A region of intense plasma wave turbulence on auroral field lines, J. Geophys. Res., 82, 1031-1050, 1977.
GURNETT, D. A., S. D. SHAWHAN, and R. R. SHAW, Auroral hiss, Z mode radiation, and auroral kilometric radiation in the polar magnetosphere: DE 1 observations, J. Geophys. Res., 88, 329-340, 1983.
HASHIMOTO, K., K. YAMAASHI, and I. KIMURA, Three-dimensional ray tracing of electrostatic cyclotron harmonic waves and Z mode electromagnetic waves in the magnetosphere, Radio Sci., 22, 579-594, 1987.
HELLIWELL, R. A. and J. P. KATSUFRAKIS, VLF wave injection into the magnetosphere from Siple Station, Antarctica, J. Geophys. Res., 79, 2511-2518, 1974.
HORITA, R. E. and H. G. JAMES, Source regions deduced from attenuation bands in VLF saucers, J. Geophys. Res., 87, 9147-9153,1982.
INAN, U. S., T. F. BELL, and D. L. CARPENTER, Explorer 45 and IMP 6 observations in the magnetosphere of injected waves from the Siple station VLF transmitter, J. Geophys. Res., 82, 1177-1187, 1977.
KIMURA, I., H. MATSUMOTO, T. MUKAI, K. HASHIMOTO, T. F. BELL, U. S. INAN, R. A. HELLIWELL, and J. P. KATSUFRAKIS, EXOS-B/Siple station VLF wave-particle interaction experiments: 1. General description and wave-particle correlations, J. Geophys. Res., 88, 282-294, 1983.
KIMURA, I., T. MATSUO, M. TSUDA, and K. YAMAUCHI, Three dimensional ray tracing of whistler mode waves in a non-Bipolar magnetosphere, J. Geomag. Geoelectr., 37, 945-956, 1985.
KINTNER, P. M., M. C. KELLEY, and F. S. MOZER, Electrostatic hydrogen cyclotron waves near one earth radius altitude in the polar magnetosphere, Geophys. Res. Lett., 5, 139-142, 1978.
MATSUO, T., I. KIMURA, and H. YAMAGISHI, ISIS-I and ISIS-II observation of emissions triggered by doppler-shifted Norway Omega signals, Mem. Natl. Inst. Polar Res., Special issue, 36, 165-180, 1985.
MOZER, F. S., C. W. CARLSON, M. K. HUDSON, R. B. TORBERT, B. PARADY, J. YATTEAU, and M. C. KELLEY, Observations of paired electrostatic shocks in the polar magnetosphere, Phys. Rev. Lett., 38, 292-295, 1977.
OKADA, T., I. NAGANO, K. HASHIMOTO, I. KIMURA, H. OYA, and A. MORIOKA, Design of a small loop antenna system for receiving waves in VLF and MF bands using a series-transformer network, Trans. IEICE, E70, 550-561, 1987.
478 I.KIMURA et al.
RAUCH, J. L. and A. ROUX, Ray tracing of ULF waves in a multicomponent magnetospheric plasma:
Consequences for the generation mechanism of ion cyclotron waves, J. Geophys. Res.,87,8191-8198,
1982.
ROUX, A., S. PERRAUT,J. L. RAUCH, C. DE VILLEDARY, G. KREMSER, A. KORTH, and D. T. YOUNG, Wave
particle interactions near ΩHe observed on board GEOS 1 and 2;2. Generation of ion cyclotron waves and
heating of He+ ions, J. Geophys. Res.,87,8174-8190,1982.
SCARF, F. L., R. W. FREDERICKS, D. A. GURNETT, and E. J. SMITH, The ISEE-C plasma wave investigation,
IEEE Traps. Geosci. Electronics, GE-16,191-195,1978.
SHAWHAN, S. D.,The use of multiple receivers to measure the wave characteristics of very low frequency noise
in space, Space Sci. Rev., 10,689-736,1970.
SONWALKAR, W. S. and U. S. INAN, Measurement of Siple transmitter signals on the DE 1 satellite: Wave
normal direction and antenna effective length, J. Geophys. Res.,91,154-164,1986.
TEMERIN, M., The polarization, frequency, and wavelengths of high-latitude turbulence, J. Geophys. Res.,83,
2609-2616,1978.
YOUNG, D. T., S. PERRAUT, A. ROUX, C. DE VILLEDARY, R. GENDRIN, A. KoRTH, G. KREMsER, and D. JONES,
Wave-particle interactions near ΩHe observed on GEOS 1 and 2;1.Propagation of ion cyclotron waves in
He+ -rich plasma, J. Ceophys, Res.,86,6755-6772,1981.