Äîęóěĺíň âç˙ň čç ęýřŕ ďîčńęîâîé ěŕřčíű. Ŕäđĺń îđčăčíŕëüíîăî äîęóěĺíňŕ : http://star.arm.ac.uk/~ambn/338jgd.ps Äŕňŕ čçěĺíĺíč˙: Tue May 30 18:35:08 2000 Äŕňŕ číäĺęńčđîâŕíč˙: Mon Oct 1 21:02:20 2012 Ęîäčđîâęŕ: Ďîčńęîâűĺ ńëîâŕ: coronal hole

LONG PERIOD OSCILLATIONS IN POLAR
PLUMES AS OBSERVED BY CDS ON SoHO
D. BANERJEE 1;2 , E. O'SHEA 3 and J.G. DOYLE 1
1 Armagh Observatory, College Hill, Armagh BT61 9DG, N. Ireland
2 Centre for Plasma Astrophysics, Katholieke Universiteit Leuven, Celestijnenlaan
200B, 3001 Heverlee, Belgium
3 Dept. of Pure & Applied Physics, Queens University Belfast, BT7 1NN, N.Ireland
dipu@wis.kuleuven.ac.be, E.Oshea@qub.ac.uk, jgd@star.arm.ac.uk
(Received ..... ; Accepted in final form .....)
Abstract.
We examine spectral time series of the transition region line O v 629 š A, observed
with the Coronal Diagnostic Spectrometer (CDS) on the SoHO spacecraft in July
1997. Both Fourier and wavelet transforms have been applied independently to the
analysis of plume oscillations in order to find the most reliable periods. The wavelet
analysis allows us to derive the duration as well as the periods of the oscillations.
Our observations indicate the presence of compressional waves with periods of 10­25
minutes. We have also detected a 11 \Sigma 1 minute periodicity in the network regions of
the north polar coronal hole. The waves are produced in short bursts with coherence
times of about 30 minutes. We interpret these oscillations as outward propagating
slow magneto­acoustic waves, which may contribute significantly to the heating of
the lower corona by compressive dissipation and which may also provide enough
energy flux for the acceleration of the fast solar wind. The data supports the idea
that the same driver is responsible for the network and plume oscillations with
the network providing the magnetic channel through which the waves propagate
upwards from the lower atmosphere to the plumes.
Key words: Sun: Polar plumes ­Ultraviolet: SoHO--Sun: Oscillations
1. Introduction
The study of polar plumes, the unipolar high density structures in
coronal holes, provides clues to the understanding of solar wind accel­
eration and coronal heating. Plumes subtend an angle of 2 degrees
relative to the Sun centre at low latitude and expand super­radially
with the coronal hole (Newkrik & Harvey 1968; DeForest et al., 1997).
More recently, DeForest & Plunkett (1999) produced images from LAS­
CO/SoHO, which clearly show polar plumes extending to altitudes of
25 solar radii or more, very close to the outer edge of the C­3 field of
view and above the likely Alf'venic point of the wind flow. With the
SUMER/SoHO instrument, polar plumes have been studied by Has­
sler et al. (1997) and Banerjee et al. (2000a). They have both reported
an anti­correlation between the intensity and line widths of the O vi

2 D. Banerjee, E. O'Shea and J.G. Doyle
1032 š A line in both plume and inter­plume regions, with detailed plume
structures been seen out to 1.5R fi .
Plume models have generally assumed a uniform cross section for
each plume and neglected time dependence (Habbal 1992; Walker et al.
1993). Karovska et al. (1994) first reported the existence of filamentary
structures within polar plumes from Skylab observation. In this paper
we investigate the temporal behaviour of polar plumes as observed in
the transition region line, O v 629 š A.
2. Observations and data reduction
The Coronal Diagnostic Spectrometer (CDS) onboard the Solar Helio­
spheric Observatory (SoHO) is a dual extreme ultraviolet spectrome­
ter, covering the wavelength range 150 to 780 š A, comprising of a normal
incidence and a grazing incidence spectrometer (Harrison et al. 1995).
The normal incidence spectrometer (NIS), whose data is the subject
of this paper, gives spectral images in two wave­bands (308--381 š A and
513--633 š A). In order to get good time resolution, we used the NIS in
a sit­and­stare mode. For the data reported here, the 4x119 arcsec slit
was used. Fig. 1 shows an image of the northern polar coronal hole
region taken with EIT/SoHO in Fe xii 195 š A at 06:15 UT on July
22, 1997, with the slit superimposed. Although CDS has the ability to
compensate for solar rotation, this was turned off since we did not want
to introduce any possible variations due to instrument movements. In
most cases no features were seen to transit the slit in the north polar
coronal hole during the observing period. Thus for our plume observa­
tions, the effects of the solar rotation are not considered to be impor­
tant.
Table I. Details of the temporal sequence CHROM N4.
Date Dataset Pointing Start End Lines Observed
X, Y UT UT
22 July 1997 s8474r00 81, 936 22:25 23:25 O v, Mg ix, Fe xvi
23 July 1997 s8474r01 80, 937 00:18 01:18 O v, Mg ix, Fe xvi
24 July 1997 s8488r00 31, 958 21:40 23:35 O v, Mg ix, Fe xvi
The data discussed here were selected from observing periods 22­24
July 1997. The observations are summarised in Table I. Three temporal
plume—oscil—v4.tex; 27/04/2000; 11:36; no v.; p.2

Long Period Oscillations in Polar Plumes 3
Figure 1. Position of the observing slit for the s8474r00 dataset (22 July 1997) on
an EIT/SoHO image in Fe xii 195 š A (courtesy of the EIT consortium).
series datasets were obtained for the three lines of O v 629 š A (log
T e =5.4 K), Mg ix 368 š A (log T e =6.0 K) and Fe xvi 335 š A (log T e =6.3
K). The counts for the Mg ix and Fe xvi lines were too low and so
these lines were not used for the power analysis. To improve the signal­
to­noise of this data we binned by three pixels along the slit (i.e. 3 \Theta
1.68 arcsec), in effect creating new pixels of 5\Theta4 arcsec 2 .
Using the standard CDS software procedure VDS CALIB, we debi­
ased and flat fielded the data. The resulting data after running this
procedure were in units of photon­events/pixel/sec and multiplying by
the exposure time yielded units of photon­events/pixel. The procedure
CDS CLEAN was used to clean the data of cosmic ray hits.
The data was checked for periods using two methods, a Fourier
analysis (see Doyle et al. 1999) and a wavelet analysis (see Torrence &
Compo 1998). Power spectra are obtained from the Fourier transforms
of the auto­covariance functions, multiplied by a window function to
reduce the variance of the noise. The power spectra are normalized
in such a way that the expected mean noise level equals 2. For the
smoothed spectra we used the Tukey window. Because the mean noise
level and its variance are known we are able to derive confidence limits
for spectral features. For intensity and velocity power spectra we use
plume—oscil—v4.tex; 27/04/2000; 11:36; no v.; p.3

4 D. Banerjee, E. O'Shea and J.G. Doyle
confidence levels of 99.9%. To determine the Doppler shifts, wavelength
calibration is needed. We use the `limb method', where we assume that
above the limb all (non­radial) wave or mass motions on average cancel
out. In the absence of radiative transfer effects, the above limb Doppler
shift must on average be zero. This method was also used by Doschek
et al. (1976) and Peter & Judge (1999). Note that the Doppler shift in
the coronal hole is independent of the rest wavelength as everything is
calibrated relative to the limb.
The localised (in time) nature of the wavelet transform allows us to
study the duration of any statistically significant oscillations as well as
their period. So to find the most reliable periods, we also performed
wavelet analysis on the data. By decomposing a time series into time­
frequency space, one is able to determine both the dominant modes of
variability and how those modes vary in time. The wavelet software uses
the Morlet wavelet and moreover allows the calculation of confidence
levels. Again we choose a confidence level of 99.9%. We should point
out that wavelet transforms suffer from edge effects at both ends of the
time series. The region in which these effects are important are defined
by the `cone of influence' (COI) (see Torrence & Compo 1998).
Figure 2. The power above the significance level as a function of frequency of
oscillation and spatial position along the slit for the s8474r00 dataset (upper panel).
The gray scale plot shows the space­time behaviour of the intensity in the O v 629 š A
line (lower panel). The gray scale coding has the most intense regions as dark. The
right panels show the counts summed over all time against the slit locations. Each
pixel is 5 arcsec with the limb at approximately pixel 18.
plume—oscil—v4.tex; 27/04/2000; 11:36; no v.; p.4

Long Period Oscillations in Polar Plumes 5
3. Results
The pointing of the slit was positioned at the solar limb in such a
way that part of the slit was outside the limb and the remainder on
the disk part of the coronal hole. One can see from the overlay of the
EIT/SoHO image and the slit location, Fig. 1, that the coronal hole
was well extended on the disk during our observing period of July
'97. Thus our slit covers network regions in the coronal hole and the
plume at the same time. Fig. 2 shows the power spectra in the 0­4
mHz range together with the contrast enhanced intensity map for the
dataset s8474r00. In the upper panel of Fig. 2, black indicates power
above the 99.9% detection level. In general we find significant power
within the range 0.5­1.75 mHz at different locations across the slit.
Figure 3. A typical network region, corresponding to pixel 5 of s8474r00 for the
O v 629 š A line (see Fig. 2), (a) & (c) shows the intensity and velocity power spectra
respectively, (b) shows the variation of intensity with time (light curve) and (d)
shows the velocity oscillation. In panels (a) & (c) the lighter line corresponds to
unsmoothed data and the bold line corresponds to the smoothed data. The solid and
the dashed horizontal lines represent 99.9% significance levels for the unsmoothed
and smoothed power respectively.
plume—oscil—v4.tex; 27/04/2000; 11:36; no v.; p.5

6 D. Banerjee, E. O'Shea and J.G. Doyle
To bring out the details of the original intensity map we have filtered
out the bright components in the image. The intensity map I(y; t) is
convolved in the time direction with a Gaussian G(t). This results in
a smoothed image S(y; t) = I \Lambda G which contains no high frequencies.
Then dividing the original intensity map by the smoothed map results
in the contrast enhanced map, i.e. C(y; t) = I(y; t)=S(y; t) (see Doyle
et al. 1999 for details). In this contrast enhanced image (lower panel of
Fig. 2), the solar north­south (SOLAR \Gamma Y) direction is in the vertical
axis, the horizontal axis is time. Fluctuations in the bright features are
clearly visible and their appearance seems periodic with a periodicity of
¸ 12 minutes. The total number of counts in a pixel (summed counts)
during the observation is shown in the right columns of Fig. 2, and
is useful in identifying the location of the solar limb and the different
network enhancements.
Below we discuss our results for the network regions in the coro­
nal hole and plumes separately and try to find whether there is any
connection between these regions.
3.1. Network oscillations
We will concentrate on pixel 5 of the s8474r00 dataset (see Fig. 2 for
the location), which corresponds to a network boundary. In Fig. 3, the
O v intensity and velocity power spectrum for this network boundary
is shown in panels (a) and (c) respectively. The lighter and the dark­
er line correspond to the unsmoothed and smoothed power spectra
respectively. The solid and dashed horizontal lines represent the 99.9%
significance level of the unsmoothed power and the smoothed power
respectively. The corresponding intensity and velocity variations with
time are plotted in panels (b) and (d) respectively. Since we are looking
for long period (low frequency) waves, a low­pass filter of everything
above 4 mHz has been applied to the velocity variations and the result
of this filtering is shown as the bold line in panel (d). The O v inten­
sity power shows a strong peak around 1.5 mHz, whereas the velocity
power shows a peak around 0.6 mHz. Note that we also find an average
redshift of 4 km s \Gamma1 in this network boundary (see Fig. 3d).
The results from the wavelet analysis for the same pixel is present­
ed in Fig. 4. The thick contours in the phase plot encloses regions of
greater than 99.9% confidence. Cross­hatched regions, on either side
indicate the `cone of influence', where edge effects become important.
The time frequency phase plane plot shows significant power between
0.8­1.7 mHz, with a strong peak around 1.5 mHz (¸ 11 min period). In
the right panel the global wavelet spectrum is plotted, which is just the
average of the wavelet power spectrum over time. The dotted line in
plume—oscil—v4.tex; 27/04/2000; 11:36; no v.; p.6

Long Period Oscillations in Polar Plumes 7
Figure 4. The wavelet analysis for O v in the network at the same location as Fig. 3.
The left panel shows the time frequency phase plot corresponding to the variations
of Fig. 3b. The right hand panel shows the average of the wavelet power spectrum
over time.
Figure 5. The wavelet analysis for O v in a typical network location (pixel no 6)
for the s8474r01 dataset. The corresponding light curve is shown at the top panel.
plume—oscil—v4.tex; 27/04/2000; 11:36; no v.; p.7

8 D. Banerjee, E. O'Shea and J.G. Doyle
the global wavelet spectrum is again the 99.9% significance level. Note
that in the wavelet time­frequency plot, the power is strong around 1.5
mHz only for the latter half of the observing sequence. The smoothed
Fourier transform (see Fig. 3) shows the same behaviour as the global
wavelet spectrum.
Choosing pixel position 6 in the s8474r01 dataset, which corresponds
to the same network boundary at a later time (¸ 40 minutes later),
Fig. 5 shows that there is significant power around 1.5 mHz, particularly
during the middle part of the temporal sequence. There is also power
at lower frequency, 0.5 mHz, which is due to edge effects.
Figure 6. A plume region, corresponding to pixel 16 of s8474r00 for the O v 629 š A
line (see Fig. 2 for location). (a) & (c) shows the intensity and velocity power spectra
respectively, (b) shows the variation of intensity with time (light curve) and (d)
shows the velocity oscillation. Representations are the same as Fig. 3.
plume—oscil—v4.tex; 27/04/2000; 11:36; no v.; p.8

Long Period Oscillations in Polar Plumes 9
Figure 7. The wavelet analysis for O v in the plume for dataset s8474r00 at pixel
position 16 of Fig. 2 (same location as Fig. 6). The left panel shows the time fre­
quency phase plot corresponding to the light curve as shown in Fig. 6b. The right
hand panel shows the average of the wavelet power spectrum over time.
Figure 8. The wavelet analysis for O v in the plume for dataset s8474r01 and at
pixel position 16.
plume—oscil—v4.tex; 27/04/2000; 11:36; no v.; p.9

10 D. Banerjee, E. O'Shea and J.G. Doyle
3.2. Plume Oscillations
Concentrate on pixel position 16 of the s8474r00 dataset, which rep­
resents a typical plume region, Fig. 6 shows the velocity and intensity
variations plus the resulting power spectra. The intensity power shows
a strong peak around 1.65 mHz, and a weaker peak around 0.4 mHz,
the velocity power peaks around 0.6 mHz. We find a substantial blue
shift when the counts are low (i.e for the first half of this time series),
but when the oscillations are clear and stronger we find less blue shifts
(last part of the time series). With the wavelet (Fig. 7), we once again
find strong power around 1.65 mHz (¸ 10 min period) for the latter half
of the observing sequence. Note also that the wavelet analysis clearly
shows that the power is significant (depicted by thick contour) for the
second half of the time series. These oscillations are also clear in the
light curve (see Fig. 6). For the s8474r01 dataset we choose the same
pixel 16 and present the power in Fig. 8. The upper panel shows the
light curve. The phase plot shows that we have significant power around
1.3 mHz (¸ 13 min period) for the first 25 minutes of the observing
sequence (it probably corresponds to the intensity jump during that
interval, see the light curve in the top panel). There is also strong pow­
er around 0.5 mHz for the entire sequence. The global spectrum shown
in the right panel also depicts the same behaviour.
Figure 9. Same as Fig. 2, for s8488r00 dataset. Power spectra across the slit in the
upper panel and space time behaviour in the lower panel.
plume—oscil—v4.tex; 27/04/2000; 11:36; no v.; p.10

Long Period Oscillations in Polar Plumes 11
Now we turn our attention to the s8488r00 dataset, which is a longer
time series, of 2 hours. Fig. 9 shows the power spectra in the 0­4 mHz
range together with the contrast enhanced intensity map. We find sig­
nificant power within the range 0.2 ­ 1 mHz at different locations across
the slit. From the contrast enhanced image, fluctuations in the bright
features seems to have a periodicity of ¸ 25 minutes. As before, the total
number of counts (summed counts) during the observation is shown in
the right column of Fig. 9, and is useful in identifying the solar limb.
Fig. 10 shows the results of the Fourier analysis and Fig. 11 shows
the results from the wavelet analysis for pixel position 13 of s8488r00
(see Fig. 9 for the location). The intensity and velocity power spectra,
Figs. 11a & c respectively, both show strong power around 0.5 mHz.
The wavelet analysis (Fig. 11) also shows a similar behaviour. Note that
the velocity wavelet shows a second peak around 1.5 mHz. On average,
the line shifts seems to be negligible. From the wavelet figures it is also
clear that the oscillations are significant for the entire duration of the
observing sequence.
Figure 10. A plume region, corresponding to pixel 13 of s8488r00 (same as Fig. 6).
(a) & (c) shows the intensity and velocity power spectra respectively, (b) shows the
variation of intensity with time (light curve) and (d) shows the velocity oscillation.
plume—oscil—v4.tex; 27/04/2000; 11:36; no v.; p.11

12 D. Banerjee, E. O'Shea and J.G. Doyle
Figure 11. The wavelet analysis for O v in the plume for dataset s8488r00 at pixel
position 13 of Fig. 9 (same location as Fig. 10). The upper panels show the intensity
power corresponding to the variations of Fig. 10b. The lower panels show the velocity
power corresponding to the variations of Fig. 10d.
4. Discussion
Polar plumes are the most prominent features in polar coronal holes
where they play an important role in the generation of the high speed
wind. SoHO observations have confirmed that plumes are denser and
cooler than the surrounding regions. From SUMER/SoHO observa­
tions, Hassler et al. (1999) showed that the plume appears wider and
more diffuse in Ne viii 770 š A, with the base of the plume originating at
plume—oscil—v4.tex; 27/04/2000; 11:36; no v.; p.12

Long Period Oscillations in Polar Plumes 13
the intersection of the network boundary. Thus it is important to study
not only the dynamics of the network regions in the coronal holes but
also the plumes which are believed to be originating from the bound­
aries of these network regions (DeForest et al. 1997).
From Fourier analysis of our network observations we find a clear
periodicity of between 10­12 minutes in intensity. The wavelet anal­
ysis indicates that the power is most significant around 1.5 mHz (11
min period) over a part of the time sequence, which suggests that the
driver is working intermittently. The nature of these oscillations may
be the same as those observed in the polar plumes (e.g. their inter­
mittent nature). If one compares Fig. 4 (the network) with Fig. 7 (the
plume), one can notice that the power is most significant for the same
part of the time sequence and also that it peaks around 1.5 mHz. This
supports the idea that the same driver is responsible for oscillations
in both regions. The network boundary probably provides the mag­
netic channel through which the waves propagate upwards, from the
lower atmosphere up to the polar plumes. Thus these waves can be
modelled as waves propagating through magnetic tubes or magnetic
slabs. Nakariakov & Roberts (1995) have presented the dispersion of
magneto­acoustic modes in a magnetic slab with a steady flow and
compared these with coronal loops. A more realistic model with coro­
nal hole parameters is badly needed to find a connection between the
network boundary and the origin of polar plumes.
Hassler et al (1999) reported strong blue shifts along network bound­
aries in a coronal hole which they attributed to the open magnetic field
structures. However, here we find red­shifts (see Fig. 3d) for a typical
network region. This difference could be simply due to the fact that
Hassler et al. were looking at the Ne viii 770 š A line, which is formed
in the high transition region and at heights where the magnetic struc­
tures are mostly open, whereas we observed O v 629 š A, formed lower
in the transition region, where the magnetic field lines are probably
closed (O'Shea et al., 2000). The implication for this is that the energy
going into heating the plasma or local acceleration depends crucially
on the local magnetic field topology. We should also remind readers
that the SUMER/SoHO spectrograph has a much better velocity res­
olution than CDS/SoHO, so our velocity measurements provide only a
qualitative view.
Now we shall concentrate on the plume regions and discuss the rel­
evance of our results with similar results obtained from other instru­
ments. High­cadence SoHO/EIT observations indicate that quasi peri­
odic fluctuations with periods of 10­15 minutes are present in polar
plumes (DeForest & Gurman 1998) with a filamentary structure with­
in the plume, on a spatial scale of 3­5 arc sec. These authors con­
plume—oscil—v4.tex; 27/04/2000; 11:36; no v.; p.13

14 D. Banerjee, E. O'Shea and J.G. Doyle
clude that the waves are either sound waves or slow magneto­acoustic
waves, propagating along the plumes at ¸ 75 -- 150 km s \Gamma1 . Using
the white light channel (WLC) of UVCS/SoHO, Ofman et al. (1997)
reported density fluctuations in coronal holes on a time scale of 9.3 \Sigma
0.4 minutes which may indicate the presence of compressional waves
further out in the corona, at 1.9 R fi . These authors, also reported
that the relative wave amplitude increase with height out to about
1.2 R fi which may indicate that slow magneto­acoustic waves propa­
gating through the plumes. Theses authors also presented a nonlinear,
two­dimensional, MHD simulation of magneto­acoustic waves in plumes
for typical coronal conditions which are consistent with observations.
Recently, Ofman et al. (2000) detected quasi periodic variations in the
polarization brightness (pB) at 1.9 R fi , at both plume and inter plume
regions. Their Fourier power spectrum shows significant peaks around
1.6­2.5 mHz and additional smaller peaks at longer and shorter time­
scales. Their wavelet analysis of the pB time series shows that the
coherence time of the fluctuations is about 30 minutes.
Our observations indicate the very clear presence of oscillations in
polar plumes, with periods of 10­25 minutes. Recently, Banerjee et al.
(2000b) reported on the existence of very long period (¸25 minute)
compressional waves in polar plumes, using a 4 hour long time series
observed with CDS in July '99. They studied the dynamics of a macro­
spicule and reported on it's effect on the background plume plasma. It
was noted that the macro­spicule was probably not connected with the
plume. In the present analysis the power spectra obtained by Fourier
and wavelet transforms have both established the existence of simi­
lar long periods. For our one hour time sequences (e.g. s8474r00 &
s8474r01) the dominant periodicity is in the 10­13 minute range, where­
as for the longer time sequence s8488r00 (¸ 2 hrs) the dominant period
is ¸ 25 minutes, although the shorter periods are still present. Even
from the contrast enhanced space time plots (see lower panel of Fig. 9)
one can clearly see a 25 minute period. The coherence time of the
1.5 mHz fluctuations in both regions (network and plume) is about
30­40 minutes (see Figs. 4, 5, 7 & 8) .
From these results it appears that if one has a shorter time series,
then the very long periods fall within the COI of the wavelet phase plot,
so it is not possible to say with any confidence if long period waves are
present. It is likely that the waves detected at 1.9 R fi by Ofman et al.
(1997, 2000) using UVCS/SoHO and the waves detected by DeForest &
Gurman (1998) around 1.2 R fi using EIT/SoHO are the same as those
reported here and observed by CDS/SoHO. These CDS observations
were taken in the polar plumes, very close to the solar limb and also in
the network boundary of the coronal holes.
plume—oscil—v4.tex; 27/04/2000; 11:36; no v.; p.14

Long Period Oscillations in Polar Plumes 15
Alf'venic oscillations are essentially velocity oscillations and do not
cause any density fluctuations. The compressional modes may however
reveal themselves in the form of intensity oscillations through a varia­
tion in the emission measure. However in their observations Ofman et
al. (1997) also observed density fluctuations. This fact, together with
the oscillations in intensity, allows us to interpret the waves as slow
magneto­acoustic in nature. Ofman et al. (1999) presented analytical
solutions for the propagation of magneto­acoustic waves in a gravita­
tionally stratified, linear, one­dimensional model of the polar plumes.
They find that, for typical coronal conditions, 10­15 minute waves are
propagating in the plume. Waves with periods longer than 70 minutes
will become evanescent. We feel that with a different set of parameters,
their calculations can also give 25 minute periods.
The localised (in time) nature of the wavelet transform has enabled
us to quantify the duration of the significant oscillations as well as
their periods. The wavelet analysis indicate that these oscillations are
not present all the time. Rather it appears that something is driv­
ing the oscillations intermittently, this driving also increases the line
intensity as we have significant power present during those intervals.
The primary cause of the 10­25 minute periodicity is not yet known.
The waves are produced in short bursts with coherence times of about
30 minutes. It appears that these waves are generated in the lower
atmosphere (below the transition region), and it may be that some
non­linear transformation of the 5 minute photospheric oscillations is
taking place. One should note that in the plume regions we have also
detected blue shifts (see Fig. 6d), which indicates outward propaga­
tion. The long period plume waves can provide significant momentum
and energy for the heating of coronal holes and the acceleration of the
solar wind. The energy carried by the slow magneto­acoustic waves can
be estimated as ae[(ffiv) 2 =2]v s
, where ffiv is the wave velocity amplitude,
and v s
= c s
= 150 km s \Gamma1 in the low fi coronal plasma. The non­
thermal velocity of the O v 629 š A line is ¸ ¸ 29 km s \Gamma1 in the `quiet
Sun' (Teriaca et al. 1999), where ¸ is related to the wave amplitude as
¸ 2 = (ffiv) 2 =2. Using ae = 1:67 \Theta 10 \Gamma15 gm cm \Gamma3 , we get a wave energy
flux of ¸ 2:1 \Theta 10 5 ergs cm \Gamma2 s \Gamma1 . This is comparable to the total ener­
gy flux required to accelerate the fast solar wind. Recent UVCS/SoHO
observations (Giordano et al. 2000) have established that the inter­
plume lanes provides the channel for the acceleration of the fast solar
wind. They have shown that in inter­plume lanes at 1.7 R fi , the corona
expands at a rate of between 105 and 150 km s \Gamma1 , that is, much faster
than in the plumes, where the outflow velocity is between 0 and 65
km s \Gamma1 . Thus it is becoming important to know whether one can find
plume—oscil—v4.tex; 27/04/2000; 11:36; no v.; p.15

16 D. Banerjee, E. O'Shea and J.G. Doyle
different type of waves in these two regions. We hope to discuss these
topics in a forthcoming paper.
Acknowledgements
Research at Armagh Observatory is grant­aided by the Dept. of Edu­
cation for N. Ireland while partial support for software and hardware
is provided by the STARLINK Project which is funded by the UK
PPARC. Wavelet software was provided by C. Torrence and G. Compo,
and is available at URL: http://paos.colorado.edu/research/wavelets/.
This work was supported by PPARC grant GR/K43315. DB wishes
to thank the ONDERZOEKSRAAD of K.U.Leuven for a fellowship
(F/99/42) and also Armagh Observatory for a short term fellowship.
We would like to thank the CDS and EIT teams at Goddard Space
Flight Center for their help in obtaining the present data.
References
Banerjee D., Teriaca L., Doyle J.G., & Lemaire P., 2000a, Solar Phys (in press)
Banerjee D., O'Shea E. , & Doyle J.G., 2000b, A&A 355, 1152
DeForest C.E., & Plunkett S. P., 1999, AAS Meeting 194, 108.01
DeForest C.E., Hoeksema J.T., Gurman J.B., Thompson B.J., Plunkett S.P.,
Howard R., Harrison R., & Hassler D.M., 1997, Solar Phys. 175, 393
DeForest C.E., & Gurman J.B., 1998, ApJ 501, L217
Doschek G.A., Feldman U., Bohlin J.D., 1976, ApJ 205, L177
Doyle J.G., van den Oord G.H.J., O'Shea E., Banerjee D., 1999, A&A 347, 335
Giordano S., Antonucci E., Noci G., Romali M., & Kohl, J.L., 2000, ApJ 531, L79
Habbal S.R., 1992, Ann. Geophys., 10, 34
Harrison et al. 1995, Solar Phys 162, 233
Hassler D.M., Wilhelm K., Lemaire P., & Sch¨uhle U., 1997, Solar Phys. 175, 375
Hassler D M., Dammasch I. E., Lemaire P., Brekke P., Curdt W., Mason H.E., Vial
Jean­Claude., Wilhelm K., 1999, Science 283, 810
Karovska M., Blundell S.F., & Habbal S., 1994, ApJ 428, 854
Nakariakov V.M., & Roberts B., 1995, Solar Phys 159, 213
Newkirk G., Jr., & Harvey J., 1968, Solar Phys 3, 321
Ofman L., Romali M., Poletto G., Noci G., & Kohl J.L., 1997, ApJ 491, L111
Ofman L., Nakariakov V.M., & DeForest C.E., 1999, ApJ 514, 441
Ofman L., Romali M., Poletto G., Noci G., & Kohl J.L., 2000, ApJ 529, 592
O'Shea E., Gallagher P.T., Mathioudakis M., Phillips K.J.H., Keenan F.P., Kat­
siyannis A.C., 2000, A&A, in press
Peter H., & Judge P. G., 1999, ApJ 522, L77
Teriaca L., Banerjee D., & Doyle J.G., 1999, A&A 349, 636
Torrence C., & Compo G.P., 1998, Bull. Amer. Meteor. Soc. 79, 61
Walker A.B.C., DeForest C.E., Hoover R.B., & Barbee T.W., 1993, Solar Phys 148,
239
plume—oscil—v4.tex; 27/04/2000; 11:36; no v.; p.16