No Title

1 - Max-Planck-Institut für Sonnensystemforschung (MPS), Max-Planck-Str.2, 37191 Katlenburg-Lindau, Germany
2 - National Astronomical Observatory of Japan (NAOJ), Mitaka, Tokyo 181-8588, Japan

Abstract
Aims. Coronal plumes, along with their structure and rôle in the acceleration of the fast solar wind, are of considerable importance in understanding the physics of the solar atmosphere, in particular, when contrasted with the adjacent plasma, the so-called interplume regions.
Methods. Observations of coronal plumes in a coronal hole were obtained with the help of spectrographs and imagers on Hinode, SOHO, STEREO, TRACE, and with ground-based instrumentation. The measurements were performed by a large international team in April 2007.
Results. First results of this study from Hinode, SOHO, and STEREO are presented here, at the level of observational facts, leaving the task of detailed analysis to the future.

1 Introduction

Polar coronal plumes seen during solar eclipses can now be studied with space-borne telescopes and spectrometers. The sources of these plumes are still not fully identified, and it is being debated whether they contribute substantially to the fast solar wind (Gabriel et al. 2003; Teriaca et al. 2003; DeForest et al. 1997; Antonucci et al. 2004; Wilhelm et al. 1998). An understanding of the processes of plume formation requires better knowledge of the physical conditions in plumes and the surrounding interplume environment. Specifically, the electron densities and temperatures, the effective ion temperatures and non-thermal motions, the plume cross-section relative to the size of the coronal hole (CH), and the plasma bulk speeds, all have to be known. Some of these aspects have motivated our work here.

2 Multi-spacecraft observations

2.1 Observations by EIS and XRT

Several coronal bright points (BP), which are thought to be the footpoints of plumes (Ahmad & Withbroe 1977; Lemaire et al. 1997; Wang et al. 1997), are seen in soft X rays by XRT and in Fe XII 19.5 nm emission by EIS. They allow an exact co-alignment. The soft X-ray image in Fig. 1 was obtained by XRT at 22:39 UTC on April 7. The plus signs indicate locations of plume footpoints determined from EUVI (see Sect. 2.2). Two footpoints around x=-80 $^{\prime \prime }$ are close to BPs. However, no prominent BP is found near another plume footpoint at about x=70 $^{\prime \prime }$ . The brightness enhancement at x=100 $^{\prime \prime }$ is behind the limb. The BP on the limb at x=-80 $^{\prime \prime }$ attained its peak radiance at that time, but underwent transient brightenings on typical time scales of 10 min, as shown in the lightcurve in Fig. 2. Some of these brightness enhancements were associated with X-ray jets. This interesting aspect is not relevant for this communication but is discussed elsewhere (e.g., Cirtain et al. 2007).

$\begin{figure} \par\includegraphics[width=8cm]{9065fig1.eps} \end{figure}$	Figure 1: Soft X-ray image obtained with the Al-poly filter of XRT. Radiances are on a logarithmic scale and plus signs in the plot mark the location of plume footpoints (cf., Fig. 4). The lightcurve of the BP marked with a box is shown in Fig. 2.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=8cm]{9065fig2.eps} \end{figure}$	Figure 2: Soft X-ray lightcurve of the bright point marked with a box in Fig. 1. A sudden burst started at 22:33 and lasted until the end of the observation, i.e. at least for 8 min.
Open with DEXTER

Figure 3 shows a radiance map in the spectral line Fe XII 19.5 nm obtained with EIS. Although this image is not a snapshot - it took 5 h to complete the raster scan - quasi-stationary plume structures are clearly seen above the limb. Two of the EUVI footpoints were covered by the scan. The plume is seen as a weak enhancement in Fe XII emission, since the formation temperature of Fe XII ( $\approx$ 1 MK) is above the plume plasma temperature. Note that the distribution of BPs is different from what is shown in Fig. 1, because of the time difference between the data. The EIS scan was run from 00:22 to 05:48 UTC many hours before the XRT image was taken. Sequences of XRT images (not shown) indicate that BPs, in addition to their short-term variability mentioned above, vary substantially on time scales of hours, while plume structures are known to be quasi-stationary and can last for one day or more.

$\begin{figure} \par\includegraphics[width=8cm]{9065fig3.eps} \end{figure}$	Figure 3: The Fe XII 19.5 nm radiance map obtained with EIS covers a range of 373 $^{\prime \prime }$ in the eastern section of Fig. 1.
Open with DEXTER

2.2 EUVI observations

Stereoscopic images (obtained on April 7 at 22:07 UTC) in the 17.1 nm passband of the EUVI instruments on the STEREO A and B spacecraft were used to reconstruct the 3D geometry of five plumes as shown in Fig. 4. The formation temperatures of the Fe IX and Fe X lines in this band ( $\le$ 1 MK) are ideal for coronal-hole observations. Linear fits outline the plume structures in epipolar view and out-of-plane distances from the footpoints up to 0.15 $R_{\odot}$ are coded by the symbol size.

$\begin{figure} \par\includegraphics[width=9cm]{9065fig4.eps} \end{figure}$	Figure 4: Results of the 3D reconstruction in epipolar projection (view onto the southern polar region). The scales are in solar radii and the out-of-plane distances of the fitted positions are coded by the symbol size. The smallest circles denote the footpoints on the disk. The viewing direction from the Earth and the visible limb are indicated.
Open with DEXTER

The reconstruction is based on displacements of identifiable structures seen in stereoscopic view as shown in Fig. 5. The principles of these method have been described in detail by Inhester (2006). The uncertainties are still quite high, because of the small difference in the aspect angles ( $3.6^\circ$ ).

$\begin{figure} \par\includegraphics[width=18cm]{9065fig5.eps} \end{figure}$	Figure 5: Images taken by the EUVI instruments on STEREO A and B as used for a 3D reconstruction of the polar coronal plumes (differences between both images cannot be detected by visual inspection and only the A image is shown here). The locations and projected directions of five selected plumes are shown. The east-west extension is 700 pixels corresponding to 800 Mm.
Open with DEXTER

2.3 SUMER observations

A raster scan with the 4 $^{\prime \prime }$ wide slit was performed in west-east direction from about $+300\hbox{$^{\prime\prime}$ }$ to $-300\hbox{$^{\prime\prime}$ }$ . The useful slit length reached from 24 Mm in height above the limb to approximately 200 Mm, above which noise contributions affected the line ratio calculations in particular. The prominent emission lines O VI at 103.2 nm and 103.8 nm, Si VIII at 144.0 nm and 144.6 nm, Mg IX at 70.6 nm and 75.0 nm, Ne VIII at 77.0 nm, Mg VIII at 77.2 nm, and H I Ly $\beta$ at 102.5 nm were included in the spectral ranges covered. The low count rates in the interplume regions of the CH required a binning of the $\approx$ 1 $^{\prime \prime }$ wide pixels by a factor of eight for some of the lines to form spatial super-pixels.

$\begin{figure} \par\includegraphics[width=12cm]{9065fig6.eps} \end{figure}$

Figure 6: SUMER observations in April 2007 in the southern CH of the Sun. The solar limb is indicated in the upper margin by a dashed line, and the radial directions from the centre of the disk at $\pm$ $12^\circ$ with respect to the negative y axis are shown as dashed-dotted line.
a) The density-sensitive Si VIII line ratio yields an estimate of the electron density, $n_{\rm e}$ . Superimposed on the density plot are, in solid lines, the projections of the 3D plume reconstructions from the EUVI observations. Plume number 4 is not within the SUMER scan range. The projections are extrapolated by dotted lines.
b) Variations in the radiance ratio R = L_77.0/L_77.2 are indicative of an elemental abundance anomaly caused by the difference in first ionization potential of high FIP element neon and low FIP element magnesium. The ratio R exhibits very clear signatures of plume and interplume regions. The combined SUMER results on plume projections are shown as solid lines for a comparison with the EUVI results.

Open with DEXTER

The density-sensitive Si VIII line pair was used to determine the electron density, $n_{\rm e}$ , averaged along the line of sight (LOS). The variation from plume to interplume regions, therefore, is underestimated. A detailed analysis of similar observations obtained in 2005 has shown that the plume density is about five times higher than that of the environment in this altitude range (Wilhelm 2006). The electron temperature, $T_{\rm e}$ , in plumes is lower than in interplume regions (cf., Wilhelm et al. 1998), a tendency also found with the Mg IX temperature-sensitive line pair (not shown) in the present data. The density is shown in Fig. 6a. The insensitivity of the Si VIII ratio to scattered radiation is discussed by Wilhelm et al. (1998) and Wilhelm (2006). It is caused by the lines being barely visible on the disk (Curdt et al. 2001) and subtracted in the corona by the standard background correction.

The contribution functions of Ne VIII and Mg VIII overlap considerably in the temperature range just below 1 MK. Variations in the radiance ratio, R = L_77.0/L_77.2 can, therefore, be interpreted, at least partially, as changes in the elemental abundances caused by the first ionization potential (FIP) effect (Young et al. 1999), since neon and magnesium are high and low FIP elements, respectively. The ratio R was plotted in Fig. 6b, and it exhibits very clear signatures of plume and interplume regions. The same spatial area as in the upper panel was covered (obtained simultaneously). The emission of Mg VIII is also very faint on the disk. Consequently, the Si VIII argument with respect to the scattered radiation applies. The Ne VIII line is very prominent in the corona, but rather weak on the disk, in particular in CHs, so that scattered radiation in the low corona is not a concern. All emission lines and the appropriate ratios have been inspected for plume structures. They allowed us to deduce composite plume projections as shown in Fig. 6b. Comparison with the radial directions from the centre of the disk at $\pm$ $12^\circ$ with respect to the negative y axis indicates the super-radial expansion of the plume ensemble.

3 Discussion and summary

We have been able to derive the 3D plume geometry from EUVI images, although still with significant uncertainties, and to link them to other observations. Extrapolating the plume structures to the base of the corona, we could associate some plume footpoints with BPs. The temporal radiance variations in BPs are not reflected in any corresponding changes in the plumes.

The spatial radiance variations of the high FIP element neon as compared to low FIP magnesium, which outline the plume pattern, indicate abundance anomalies, although a detailed consideration of the contribution functions has still to be done.

We applied plasma diagnostics based on emission line ratios to derive the electron density, confirming that plumes have higher densities than their environment. These observationsthereby favour the view that the cross-sections of plumes are restricted along the LOS direction, and thus the total area occupied by all plumes is small compared to the size of the CH. This suggests that the plume contribution to the fast solar wind is relatively small.

There are plans for further observations with a much better stereoscopic view later in 2007 and 2008. An International Space Science Institute (ISSI) study Team lead by one of us aims to continue and extend the work presented here, motivated by the prospect of more and better observations.