Astron. Astrophys. 342, 867-880 (1999)

7. Dopplergrams of H dark features

In this section the link between some observed H dark features and their associated observed Dopplershifts is investigated. In order to do so we present a zoomed portion of the filament channel observed at 08:43 UT (Fig. 5).

7.1. Method for measuring dopplershifts

The Doppler velocities presented in Fig. 5c have been calculated by using the method of line-bisector at Å from line center. They correspond to line-of-sight velocities only in the case of optically thick structures. For typical chromospheric fine structures, the expected underestimation factor is between 1 and 4 (Alissandrakis et al., 1990). Radiative transfer calculations using "cloud models" are necessary to convert dopplershifts into accurate velocities.

The zero-value of dopplershifts is determined by averaging the data over the whole field of view. Blueshifts of filaments are included as well as redshifts of network cell boundaries, so that the accuracy of the result is probably not better than a few hundreds of m s^-1. However, relative dopplershifts between neighbouring points are determined more accurately.

The minimum and maximum dopplershifts in the field of view are 7 km s^-1. The filament body shows typical weak upward motions from few hundred m s^-1 up to 3 km s^-1, in accordance with several previous studies (e.g. Schmieder, 1990 and Mein et al., 1994).

7.2. Description of fine structures

Four boxes are defined on Fig. 5a-d, which enclose the fine structures discussed in the following. Box 1 shows the forming top-left foot F2 of the filament, Box 2 shows a small lateral foot on the right of the filament which disappears later, Box 3 shows an elongated and curved fine structure, and Box 4 shows S4 (as defined in Fig. 1a) which is the "M-shaped" group of fibrils discussed in Sect. 6.2.3.

Comparing Fig. 5b and Fig. 5d confirms that the lateral dip structures are created by parasitic magnetic polarities, i.e. positive ones for Boxes 1,3,4, and negative ones for Box 2. Comparing the shape of the H features (in Fig. 5a) with the distribution of dips (in Fig. 5b), it is noteworthy that the orientation of the fine structures do not necessarily correspond to the orientation of their inner magnetic field. For example, looking at Box 3 clearly shows that is perpendicular to the elongated fine structure, and that this elongated shape is mainly due to the continuous dips pattern.

7.3. Vertical motions in the fine structures

The observed filament was nearly at the disc center on September 25 (S2, E5), consequently the Doppler velocities nearly correspond to the vertical velocities of the plasma.

Let us focus on the four boxes. A first look at Fig. 5c shows that their associated H dark fine structures mainly present low upward velocities (compare Figs. 5a,c). By averaging typical dopplershifts taken at 20 points along each structure, we obtain the values listed in Table 3.

Table 3. Average Doppler velocities measured in the dark H fine structures which are located in the four Boxes defined on Fig. 5.

The observed velocities in the fine structures are much closer to the driving velocities of photospheric magnetic polarities (0.1 km s^-1 1 km s^-1) than to the typical magnetic velocities (i.e. Alfvén velocity), in spite of their possible under-estimation due to their small optical thickness (see discussion in Sect. 7.1). This has the following natural explanation. The slow horizontal motions of parasitic polarities are expected to cause a global quasi-static evolution of the 3-D magnetic configuration: the magnetic field lines are progressively deformed so that the dips are horizontally and vertically displaced. Because the ionisation is sufficient to freeze the plasma in the magnetic field, we propose here that the Dopplershifts represent these vertical motions. The observed motions are in accordance with the support of the plasma in magnetic dips (in the vertical direction the tension term of the Lorentz force nearly balances the gravity, plasma and magnetic pressure forces). Along the magnetic field, so in a nearly horizontal direction, plasma flows can be driven, for example by pressure differences. Observations show typically higher horizontal velocities than vertical ones by a factor 10. (e.g. see Schmieder, 1990). Such horizontal velocities can be at the origin of a shift between the observed H dark material and the computed dips. The present results put a constraint on this effect.

© European Southern Observatory (ESO) 1999

Online publication: February 23, 1999
helpdesk.link@springer.de