Solar wind acceleration in low density regions
L. Teriaca, G. Poletto , M. Romoli† and D. Biesecker
†
£
Osservatorio Astrofisico di Arcetri, Largo E. Fermi 5, 50125 Firenze, Italy
Dipartimento di Astronomia e Scienza dello Spazio, Università di Firenze, Largo Fermi 5, 50125 Firenze, Italy
££
Emergent I. T., Inc., NASA/GSFC, Greenbelt, USA
Abstract.
High speed solar wind is known to originate in polar coronal holes which, however, are made up of two
components: bright, high density regions known as plumes, and dark, weakly emitting low density regions known
as interplumes. Recent space observations have shown that the width of UV lines is larger in interplume regions
[see e. g. 1, 2] while observations of the ratio of the O VI doublet lines at 1032 and 1037 Å, at the altitude of
1.7 solar radii, suggest higher outflows in interplume regions than in plumes [3]. These results seem to locate the
source of the fast solar wind in the interplume regions.
The present work aims at identifying the outflow speed vs. altitude profile of the O VI ions, at heights up to 2
solar radii, both in plumes and interplume regions. To this end, we examined SUMER and UVCS data taken in
the North polar coronal hole on June 3, 1996 over the altitude range between 1 and 2 solar radii.
A Doppler dimming analysis applied to our data allows us to determine the outflow speed in interplume regions
throughout the range covered by the observations. Our results favor interplumes as sources of fast wind. However,
models mimicking observations in plume regions will also be discussed.
INTRODUCTION
OBSERVATIONS
Polar coronal holes have long been recognized to be the
sources of the high speed solar wind [4], but only recently some light has been shed on where, within coronal holes, the solar wind originates. UVCS and SUMER
measurements of line widths and of the ratio of the O VI
doublet lines at 1032 and 1037 Å in the low corona (at
heliocentric distances lower than 2.5 R ) seem to indicate that the low density background plasma, rather than
the high density plume plasma, is the site where high
speed wind originates. Recent analyses of SUMER and
UVCS data [see e. g. 3, 2, 5] have shown that the width
of UV lines is larger in interplume than in plume regions,
hinting to interplumes as the site where energy is preferentially deposited and, possibly, fast wind emanates.
Moreover, the analysis of fast polar wind performed by
[6] seems to favor the low density regions as sources of
the fast wind streams, while no evidence of outflow motions in bright points and plumes observed within a coronal hole in the Ne VIII 770 Å line has been found by [7].
In this paper we present strong evidence supporting
the idea that interplume regions are sources of the fast
solar wind. The outflow velocity profile for O VI ions
in interplume areas will be given together with a model
reproducing plume observations.
The observations discussed here were acquired using
several instruments aboard SoHO on 3 June 1996 and
are shown in Fig. 1. SUMER observations comprise two
rasters of the North Polar Coronal Hole (NPCH) obtained
using the 4 300 slit and covering a total area of
280 470 (for further details [see 2]). These rasters
were aimed to obtain high signal-to-noise O VI 1032 and
1037 Å line profiles in the plume and interplume regions
above the limb out to 1.5 R . Seven UVCS spectra
comprising the O VI doublet and the Hydrogen Ly al pha
were acquired with the slit tangent to the solar limb at
altitudes up to 2.1 solar radii. A CDS raster scan of the
solar limb in the Mg IX 368 Å line provides informations
on the roots of the structures onto the solar disk.
Polar plumes are linear structures that are apparent
over the solar poles in visible light, in extreme ultraviolet
and in soft X-rays [see 8, and references therein]. After
selecting a plume and an interplume region, binning
was applied to SUMER and UVCS data to increase
adequately the S:N ratio. In such a way, line profiles were
determined at 19 different locations above the solar limb
for both plume and interplume areas.
CP679, Solar Wind Ten: Proceedings of the Tenth International Solar Wind Conference,
edited by M. Velli, R. Bruno, and F. Malara
© 2003 American Institute of Physics 0-7354-0148-9/03/$20.00
327
DATA ANALYSIS
SUMER spectral profiles were carefully fitted in order
to obtain C II 1037, O VI 1032 and O VI 1037 Å line
intensities. Despite the very high quality of the telescope
mirror, the level of stray light is not negligible when
observations of lines that are bright on disk are carried
out above the limb [9]. The level of stray light in SUMER
spectra was determined using the C II 1037 Å line as
described in [2]. Stray light in UVCS spectra was also
estimated and removed.
Errors on line intensities were calculated through Poissonian statistics and, finally, their propagation in the O VI
1032/1037 line ratio was evaluated.
RESULTS AND DISCUSSION
FIGURE 1. June 3rd 1996 map of the north polar coronal
hole as obtained using four of the instruments aboard SoHO.
Strong plumes and inter plume lanes can be identified in this
diagram out to 2.0 R¬ above the limb. An EIT image in the
Fe XII 195 Å line provides the disk image showing a well
developed coronal hole. A CDS raster in the Mg IX 368 Å
line shows the root of a large plume structure. A SUMER scan
in the O VI 1032 Å line shows how the plume develops with
altitude up to 1.5 solar radii. Above this altitude the structure of
the coronal hole is obtained throughout seven spectra acquired
in the O VI 1032 Å line with the UVCS slit normal to the
solar radius. The contrast of the SUMER and UVCS maps was
enhanced by dividing each map for its average intensity profile
in the y-direction.
O VI 1032 and 1037 Å spectral lines arise from transitions well described by a two level atomic model where,
at coronal conditions, the upper level is populated by
electron impact excitation (collisional component) and
by resonant absorption of the O VI radiation from the
transition region (radiative component),
Iobs
A
IRad 0 85 ∆E
4π O H
qTeRTe Ne2 BI DTO W RTeNe IColl
(1)
where ∆E is the energy of the transition, A OH is the
oxygen abundance relative to hydrogen, qT e the collisional excitation rate coefficient, B the Einstein absorption coefficient, I the O VI disk-averaged line intensity,
DTO W accounts for Doppler dimming and geometrical dilution factors and is function of the oxygen temper
ature TO (consisting of the components parallel To and
perpendicular To to the direction of the magnetic field
lines) and the wind velocity W . Ne is the electron number density, RTe is the oxygen ionic fraction calculated
in ionization equilibrium and 0.85 is the value of the hydrogen to electron number density ratio for a fully ionized plasma with composition given by [10]. The quantities in brackets ... are integrated along the line of sight.
We are interested in obtaining the O VI outflow profile
through the comparison of observed O VI line intensities
and ratio with values computed using Eq. 1. Both radiative and collisional components are strongly dependent
on the electron density (N e ), electron temperature (T e ),
and oxygen elemental abundance (A OH ), while the radiative component depends also on the adopted values of
I , To , To and W .
I was evaluated assuming a 1cosθ center-to-limb
line intensity variation [11] and adopting an intensity at
disk centre of 280 mW m 2 St1 for the O VI 1032 Å
line. The above value was obtained from a disk centre
spectrum obtained on June 4th 1996.
In the case of the observations here discussed, we can
assume the magnetic field lines to be perpendicular to the
line of sight. This allow us to identify To with the effective temperature Te f f associated with the Doppler width
(in km s1 ) of the observed spectral line. Doppler width
in both plume and interplume regions were obtained for
the SUMER data of this dataset by [2]. UVCS line widths
were obtained from our data.
IRad is not a strong function of To and, due to the small
difference between measured line widths in plume and
interplume, a unique Doppler width profile was assumed
as representative of both regions. Both cases of To equal
to Te (anisotropic case) and To equal to To (isotropic
case) have been explored.
Spectroscopically derived Te values never exceed 10 6
K in the range 1.05–1.3 R . [5] provide the only measurements of Te in plume and interplume regions below 1.2 solar radii while, at higher altitudes, an upper
328
FIGURE 2. O VI 1032 to O VI 1037 intensity line ratio as a function of altitude in polar coronal holes. Values obtained in
interplume regions: SUMER data (open circles), UVCS data (open squares). Values obtained in plumes: SUMER data filled circles,
UVCS data (filled squares). Solid lines represent models for interplume conditions in the anisotropic (thin lines) and partially
isotropic case (thick lines), respectively. Dashed lines represent a mixed model comprising plume and interplume lanes.
limit to the electron temperature is given by the value of
1.075 MK given by [12] from in situ measurements of
the O VII/O VI ion ratio in fast wind streams. For this
work we adopted the T e values given by [5] below 1.2
solar radii and we assumed constant temperature above
this altitude.
The O VI 1032/1037 line ratio is insensitive to T e [14]
and does not depend on the elemental abundance. It is,
however, strongly dependent on the electron density N e
as well as on the wind speed W . In order to reproduce the
observed line ratio, the N e profile as a function of height
needs to be known, together with the exciting transition
region radiation. Above 1.5 solar radii average electron
densities were obtained from the UVCS WLC and the
LASCO C2 coronograph for the day of our observations.
Densities in coronal holes have been evaluated by several authors [e. g. 15, 16]. However, in the present analysis only data obtained by [17, 5, 18, 19, 15] in 1996 (i. e.
closer in time to our data) were used to integrate our measurements. In particular, the behaviour at lower altitudes
was evaluated from data published by [5] for both plume
and interplume regions.
Using those electron density and temperature profiles
the O VI 1032 and 1037 observed line intensity ratios
below 1.2 solar radii were reproduced throughout Eq. 1
assuming no outflow speed. In these conditions (W 0),
O VI line intensities depend on the elemental abundance
only (once the density and temperature profiles have
been fixed). We found that observed line intensities be-
low 1.2 solar radii could be reproduced provided an oxygen abundance of 8.5 is adopted.
Using the adopted interplume T e and Ne profiles and
an oxygen abundance of 8.5, we found the outflow
profiles able of reproducing our interplume data (see
Fig. 2) for both isotropic and anisotropic conditions.
Above 1.3 solar radii, unless wind is accelerated and
lines are Doppler dimmed, we would be unable to reproduce UVCS data (see Fig. 3).
A realistic attempt to model our plumes observations
requires a combination of plume and interplume emissivities along the line of sight. A mixed plume-interplume
model has been, hence, created assuming a single plume
with no outflowing plasma embedded in the outflowing
interplume plasma. The line of sight fraction occupied by
the plume was estimated from Fig. 1. This model is able
to reproduce our plume observations (see Fig. 2) supporting the hypothesis that interplumes are the regions from
where fast wind streams emanate.
CONCLUSIONS
Our interpretation of the measurements of the O VI
1032/1037 line intensity ratio shows that negligible outflow velocities are consistent with the values measured
below 1.3 R , while the wind starts being accelerated
above this altitude. This is the first time that the O VI
outflow profile is provided below 1.5 solar radii: results
329
FIGURE 3. O VI ions outflow speed as a function of altitude in polar coronal holes. Values obtained in interplume for the
anisotropic case are shown as open circles while values obtained in the partially isotropic case are represented using filled circles.
As a reference, the outflow profiles given by [13] for an average coronal hole are also shown.
presented here show that the wind acceleration initiates
at lower altitudes than usually assumed ([see e.g. 13]).
8.
9.
10.
ACKNOWLEDGMENTS
11.
L. T. and G. P. are partially supported by MURST and
ASI. SOHO is a mission of international cooperation
between ESA and NASA.
12.
REFERENCES
1.
2.
3.
4.
5.
6.
7.
13.
Giordano, S., Antonucci, E., Benna, C., and others,
“Plume and interplume regions and solar wind
acceleration in polar coronal holes between 1.5 and
3.5 R¬ ”, in Fifth SOHO Workshop: The Corona and Solar
Wind Near Minimum Activity, edited by A. Wilson, ESA
SP 404, ESA Publications Division, ESTEC, Noordwijk,
The Netherlands, 1997, pp. 413–416.
Banerjee, D., Teriaca, L., Doyle, J. G., and Lemaire, P.,
Solar Physics, 194, 43–58 (2000).
Giordano, S., Antonucci, E., Noci, G., and others,
Astrophysical Journal, 531, L79–L82 (2000).
Krieger, A. S., Timothy, A. F., and Roelof, E. C., Solar
Physics, 29, 505–525 (1973).
Wilhelm, K., Marsch, E., Dwivedi, B. N., and others,
Astrophysical Journal, 500, 1023–1038 (1998).
Antonucci, E., Dodero, M. A., and Giordano, S., Solar
Physics, 197, 115–134 (2000).
Wilhelm, K., Dammasch, I. E., Marsch, E., and Hassler,
D. M., Astronomy and Astrophysics, 353, 749–756 (2000).
14.
15.
16.
17.
18.
19.
330
DeForest, C. E., Hoeksema, J. T., Gurman, J. B., and
others, Solar Physics, 175, 393–410 (1997).
Lemaire, P., Wilhelm, K., Curdt, W., and others, Solar
Physics, 170, 105–122 (1997).
Feldman, U., Mandelbaum, P., Seely, J. F., and others,
Astrophysical Journal Supplement, 81, 387–408 (1992).
Wilhelm, K., Lemaire, P., Dammasch, I. E., and others,
Astronomy and Astrophysics, 334, 685–702 (1998).
Wimmer-Schweingruber, R., Von-Steiger, R., Geiss, J.,
and others, “O5 in high speed solar wind streams:
SWICS/Ulysses results”, in Solar Composition and its
Evolution - From Core to Corona, edited by C. Frohlich,
M. C. E. Huber, and S. K. Solanki, Kluwer Academic
Press, 1998, pp. 387–396.
Cranmer, S. R., Kohl, J. L., Noci, G., and others,
Astrophysical Journal, 511, 481–501 (1999).
Li, X., Habbal, S. R., Kohl, J. L., and Noci, G.,
Astrophysical Journal, 501, L133–L137 (1998).
Kohl, J. L., Noci, G., Antonucci, E., and others,
Astrophysical Journal, 501, L127–L131 (1998).
Fisher, R., and Guhathakurta, M., Astrophysical Journal,
447, L139–L142 (1995).
Doyle, J. G., Teriaca, L., and Banerjee, D., Astronomy and
Astrophysics, 349, 956–960 (1999).
Zangrilli, L., Poletto, G., Nicolosi, P., and others,
Astrophysical Journal, 574, in press (2002).
Lamy, P., Quemerais, E., Liebaria, A., and others,
“Electronic Densities in Coronal Holes from LASCO-C2
Images”, in Fifth SOHO Workshop: The Corona and Solar
Wind Near Minimum Activity, edited by A. Wilson, ESA
SP 404, ESA Publications Division, ESTEC, Noordwijk,
The Netherlands, 1997, pp. 491–494.