Geoscience Reference
In-Depth Information
5.3.2
Observations of the IAR Spectra
As we have noted above, the SRS of IAR is mainly evident during the night
time irrespective of season, at low (e.g., see Bösinger et al. 2002 ), middle (e.g.,
see Hebden et al. 2005 ), and even high (e.g., see Semenova and Yahnin 2008 )
latitudes. Now we first consider the IAR signature detected at the mid-latitudes at a
remote site near Karimshino station (52:94 ı N, 158:25 ı E, L D 2:1) in Kamchatka
peninsula (Fedorov et al. 2006 ; Surkov et al. 2006 ). The reader is referred to
the work by Uyeda et al. ( 2002 ) for details about the equipment of the Russian-
Japanese geophysical observatory in Kamchatka. The ground-based observations
at this point have shown that the IAR signature predominantly occurred at local
nighttime. Analysis of data obtained at Karimshino station has demonstrated that
some spectrograms should be interpreted as impulse IAR excitation rather than
permanent one. To illustrate this, we have chosen a representative time interval
from 21 h till 22 h (local time) on 13 September 2000. During this interval, half-
minute samplings were used to analyze the spectra of the ULF geomagnetic
variations. Typical half-minute recordings, power spectrum densities, and dynamic
spectrograms of the magnetic variations are shown in Figs. 5.6 , 5.7 , 5.8 . The upper
panels marked H SR and D SR display the time-dependence of D (the magnetic
declination) and H (the horizontal component) components of magnetic variations
in pT in the frequency ranges of 6-20 Hz. In the middle row of the panels, the data
displayed in the top row are low pass filtered 0.25-4.0 Hz. The time in seconds
is referenced to 21 h. For example, the 30 s interval shown in Fig. 5.6 begins at
21:31:30. The function log 10Œp.B/ (B in pT) is shown in Figs. 5.6 , 5.7 , 5.8 in
the second and fourth panels in the first two rows, where p.B/ is the amplitude
probability distribution of B. The power spectrum densities of time-derivative of
the magnetic variations are shown in the second and fourth panels in the third row.
Morlet wavelet decomposition (Mallat 1999 ) was used in order to obtain the
normalized dynamic spectrograms of both components that are marked H and D at
the lower row of the panels. The vertical axis on these panels corresponds to the
wavelet center frequency. What attracts our first attention is that the enhancement of
the dynamic spectrograms occurs just at the moment of the sharp impulses, which
occasionally happen at the middle panels. For example, it is obvious from Fig. 5.6
that the impulse occurrence at the low-frequency channel (0:25-4 Hz) at the moment
t 1;914 s is accompanied by an enhancement of the dynamic spectrograms of
both components. It should be noted that the SRS builds up as a result of this impulse
and exhibits typical frequencies close to the IAR eigenfrequencies. Similarly, two
impulses seen on the second line in Fig. 5.7 cause the distinct SRSs of the dynamic
spectrograms. It is worth mentioning that the pattern of the dynamic spectrograms
are consistent with the half-minute power spectra displayed in Figs. 5.6 , 5.7 , 5.8
at the lower row. During a quiet period shown in Fig. 5.8 the SRS is not so distinct
except for the low-frequency domain of the spectrograms whereas the SRS is clearly
seen in the power spectra. The same resonance structure with distinct IAR signature
Search WWH ::




Custom Search