line-of-sight velocity fluctuations of a quiescent prominence

LINE-OF-SIGHT VELOCITY FLUCTUATIONS OF A QUIESCENTPROMINENCE

V. N. DERMENDJIEV1,∗, N. I. PETROV1, M. Tz. DETCHEV1, B. ROMPOLT2 andP. RUDAWY2

1Institute of Astronomy, 72 Tsarigradsko Chaussee, Sofia 1784, Bulgaria(E-mail: [email protected], [email protected])

2Astronomical Institute of Wroclaw University, ul. Kopernika 11, 51-622 Wrocław, Poland(E-mail: [email protected], [email protected])

(Received 13 July 2000; accepted 11 April 2001)

Abstract. Series of Hα spectra and slit-jaw Hα filtergrams of a quiescent prominence taken at Pic duMidi Observatory on 7 November 1977 are studied. The observations have been digitized by means ofan automated Joyce Loebl microdensitometer. The image processing of the Hα filtergrams reveals aninternal structure of the prominence consisting of several arches. Series of high-resolution Hα spectraobtained with the slit position located on a selected part of one of the prominence arches have beenchosen for Doppler-shift analysis. The obtained time series of the line-of-sight velocity reveal largevelocity variations near the periphery of the arch and a strong dependence of the velocity (in signand magnitude) on the position along the slit. The prominence arch shows also cyclic displacementsalong the line-of-sight direction implying Alfvén string-mode oscillations.

1. Introduction

It is well known (Tandberg-Hanssen, 1995) that high-resolution limb observationsof quiescent prominences (QPs) often reveal fine-scale filamentary structures ex-hibiting systematic, transverse to the line of sight, directed downward motion withshort time scales. When such prominences are observed on the disk as filamentsthe Doppler-shift measurements indicate both upward and downward line-of-sightdirected plasma motion, the average velocities being mainly upward. Internal mo-tions in QPs have long been under study but there is still little knowledge abouttheir origin, behaviour, and range of velocities. Pettit (1932) had found, from timesequences of filtergrams, velocities of 5 to 10 km s−1. Newton (1934) had studiedthe radial velocities of some dark disk filaments and had obtained velocities smallerthan 4 km s−1. Using the Hα high-resolution emission line in QPs single knots, TenBruggencate and Elste (1958) had obtained internal motion with velocities up to10 km s−1; in fainter QP microstructures, velocities up to 40 km s−1 were observed(Severny and Khokhlova, 1953).

Extensive measurements of line-of-sight (l.o.s.) velocities in prominences havebeen carried out by Liszka (1970), Giogolashvili and Zhugzhda (1982), Engvold

∗Deceased.

Solar Physics 202: 99–107, 2001.© 2001 Kluwer Academic Publishers. Printed in the Netherlands.

100 V N. DERMENDJIEV ET AL.

(1972), Engvold, Wiehr, and Wittman (1980), Yi, Engvold, and Keil (1991), Yiand Engvold (1991). The Hα high-resolution observations of Yi (1992) have givenvelocities of about ±1 km s−1. Recently the re-examination of some earlier high-resolution observations of limb prominences by means of Local Cross-CorrelationTechniques (Zirker, Engvold, and Yi, 1994) and with a method for calculatingthe geometric distortion between subsequent images (Pojoga and Molowny-Horas,1999) reveals flow velocities and directions that are in good agreement with thoseobtained by Doppler measurements of disk filaments. The impression is (Zirker,Engvold, and Yi, 1994) that the matter in the fine structure of prominences flowsboth upwards and downwards under forces other than gravity.

There are many observations implying oscillations in QPs, and cyclic velocityand intensity changes are commonplace (Tandberg-Hanssen, 1995). The oscilla-tions have a large range of periods, of the order of an hour to 3–5 min. High-resolution observations of quiescent filaments (QFs) show oscillations that arestrongly tied to their fine threads (Yi, Engvold, and Keil, 1991). Recent studieshave detected pronounced oscillatory variation in the Doppler velocities of QFs (Yi,Engvold, and Keil, 1991; Yi and Engvold, 1991; Tomson and Schmieder, 1991).The oscillations with periods between 12 and 16 min have been found (Yi, 1992)to be much stronger in QFs than in the nearby chromosphere.

There are two main points of view concerning the nature of the QP internalmotion. The first one, suggested by Malville (1968), supposes a flux of mechan-ical energy entering the prominence in the form of Alfvén waves. This suppo-sition stimulated many observational studies for detection of pronounced oscil-latory variations in the l.o.s. velocity of QFs (Yi, Engvold, and Keil, 1991; Yiand Engvold, 1991; Tomson and Schmieder, 1991) and QPs (Gigolashvili andZhugzhda, 1982; Bashkirtsev, Kobanov, and Mashnich, 1983; Wiehr, Stellmacher,and Balthasar, 1984; Balthasar, Stellmacher, and Wiehr, 1988; Molowny-Horaset al., 1997; Petrov, Dermendjiev, and Rompolt, 1998). The observational resultsstimulated, on the other hand, theoretical attempts aiming to explain the reportedquasi-periodical oscillations in the Doppler-shift time series in terms of propagat-ing magnetohydrodynamic (MHD) waves (Oliver et al., 1993; Oliver and Ballester,1996), and Alfvén waves (Solov’ev, 1985; Tomson and Schmieder, 1991; Jensen,Yi, and Engvold, 1994).

According to a second point of view, proposed by Jensen (1982), the observedl.o.s. velocity distribution functions imply MHD-turbulence, as is described byKraichnan (1965), and that a characteristic relationship should exist between ve-locity and eddies size.

The aim of the present work is to study the l.o.s. velocity fluctuations in aselected arch of a QP on the base of a series of high-resolution Hα spectra.

LINE-OF-SIGHT VELOCITY FLUCTUATIONS 101

Figure 1. (a) Hα filtergrams of the prominence. The line gives the position of the slit. (b) Series ofHα spectra used for Doppler-shift measurements (see the text).

2. The Observations and Method of Processing

We examine the structure and the l.o.s. velocity variations of a QP on high-resolutionHα spectra and filtergrams obtained by one of the authors (BR) on 7 November1977 at Pic du Midi Observatory. The prominence was located at the NE limband, according to the Meudon synoptic map for rotation N1681, the centre ofthe respective filament (N38 E97) was situated 7◦ behind the limb on the day ofobservations and its western nearly straight lined part has crossed the limb underan angle of about 45◦. The filament had lasted for two solar rotations and duringthe observations it had been on its first rotation.

The series of high-resolution Hα spectra have been recorded on photographicfilm using the 11-m solar horizontal telescope equipped with a 9-m spectrograph(Mouradian and Leroy, 1977) and a spectrohelioscan developed at the Astronomi-cal Institute of the Wrocław University. A slit-jaw unit coupled with an Hα camerahave provided Hα filtergrams together with the actual slit position against theprominence body. The spectra from the QP have been taken with an exposure timeof 5 s and 0.75 mm slit width.

The Hα filtergrams have revealed that the QP consists of several arches. Forthe purpose of this study we took seventeen Hα spectra obtained at a slit positionlocated on one of the arches, marked by a line in Figure 1(a). The spectra, shown inFigure 1(b), have been obtained in a time interval of about 57 min (from 10:07:48


to 11:04:35 UT) with a nearly equal time step of 3.5 min between the consecutivescans.

The selected spectra and filtergrams have been digitized by an automated JoyceLoebl microdesitometer at the National Astronomical Observatory ‘Rozhen’. Us-ing computer facilities the position of the centre of gravity of each Hα profilehas been determined with respect to the mean Hα profile of the studied volumeof the arch. The deviation of the determined wavelength in respect to the meanHα position has been interpreted in terms of microscopic l.o.s. velocity fluctua-tion; the velocity has been defined as positive for red-shifted Hα features. TheHα profiles have been plotted by means of a computer-operated device in orderto eliminate noisy and erroneous microdesitometer tracings. Finally, 816 veloc-ity measurements with particularly good signal-to-noise ratio have been obtainedtogether with other quantities required for further analysis.

3. Results and Discussion

The dependence of the l.o.s. velocity fluctuations on time and position along the slitis shown in Figure 2. The position along the slit is given in 50 µ (one pixel) stepsequal to 343 km. The corresponding velocity distribution functions along the slitare shown in Figure 3. As can be seen in these figures, the sign and the magnitudeof the l.o.s. velocity are changing along the slit, and with time, especially nearthe periphery of the arch. At the beginning of the observations the velocity at theupper part of the studied prominence volume is negative, i.e., the motion is directedto the observer, while at the lower part it becomes positive, i.e., the motion is inthe opposite direction to the observer. The maximum magnitude of the velocity isabout ±1.5 km s−1. After 31.5 min a tendency appears for the sign of the l.o.s.velocity to become positive near the upper part and negative at the lower part ofthe studied prominence volume.

The distribution function of all l.o.s. velocity fluctuations obtained (Figure 4)has a mean value of V = −0.0396 km s−1, mean square deviation of S = 0.6351and has a significant asymmetry for positive and negative velocities. There are,as well, two small peaks at both tails of the distribution. In the same figure anapproximation with a Gaussian distribution with mean value V = 0.0796 km s−1

and σ = 0.3362 is presented.The distribution function of the absolute values of the l.o.s. velocity fluctuations

(Figure 5) has V = 0.4723 km s−1 and S = 0.4264. A small peak at the tail ofthe distribution is seen, as well. Dividing the velocity data in two samples, one for|V | = 1.25 km s−1 and the other for |V | > 1.25 km s−1, we obtained distributionfunctions with V = 0.0372 km s−1 and S = 0.3046, and with |V | = 1.443 km s−1

and S = 0.1032, correspondingly.In order to trace cyclic displacement or shaking of the prominence arch along

the l.o.s. direction, which is parallel to the solar surface, we compute, for the se-


Figure 2. Dependence of the l.o.s. velocity fluctuations on the time and the position along the slit.The velocity changes sign along the slit and becomes large near the periphery of the arch.

lected prominence volume, the position of the centre of gravity of each Hα profilewith respect to the corresponding chromospheric Hα. As a representative statisticalquantity for such a displacement we take the median of the velocity distributionalong the slit.

The time dependence of the median Me(ν) of the l.o.s. velocity distributionalong the slit is shown in Figure 6, where the two-side confidence intervals anda polynomial fit are shown, as well. It is well seen that the observed volumeof the arch shakes first away from the observer (positive values of Me(ν)), thentowards the observer (negative values) and later there seems to again be a ten-dency to reverse. This implies a cyclic displacement or vibration of the prominencearch. Unfortunately, the studied series of Hα spectra is not long enough to revealthe period. In any case it seems longer than the time interval of the registration,i.e., 57 min, which suggests Alfvén string-mode oscillations studied recently byJoarder, Nakariakov, and Roberts (1997) for a model prominence fibril.

The mean value of the difference of consecutive median’s values, �Me(ν) =0.266 km s−1, can be used as an estimate of the displacement ξ = �Me(ν) =55.8 km (for an individual time step of 3.5 min).

4. Conclusion

The studied series of high-resolution Hα spectra reveals an interesting behavior ofthe l.o.s. velocity fluctuations along the slit and in the time. The distribution func-tion of the obtained 816 estimates of the l.o.s. velocity shows significant asymmetryfor positive and negative velocities, and small peaks at both tails corresponding tothe highest velocities. The series of the distribution function of the l.o.s. velocityfluctuations and the quasi-three-dimensional presentation of the velocity fluctua-tions as a function of the slit position and time suggest a quasi-cyclic change of


Figure 3. Series of distribution functions of the obtained l.o.s. velocity fluctuations.


Figure 4. Distribution function of the l.o.s. velocity fluctuations obtained from the whole series ofHα spectra.

Figure 5. Distribution function of the absolute values of the l.o.s. velocity fluctuations obtained fromthe whole series of Hα spectra.


Figure 6. Time dependence of the median Me(ν) (filled circles) of the l.o.s. velocity distribution alongthe slit. The vertical lines and the continuous curve represent the two-side confidence intervals anda polynomial fit, respectively. The observed volume of the arch shakes first away from the observer,than reverses direction with a tendency to reverse again.

the velocity magnitude and the sign, and that higher velocities are seen near theperiphery of the arch.

The Doppler-shift measured with respect to the averaged Hα profile of thechromosphere indicates cyclic displacement or vibration of the arch that could bereferred to Alfvén string modes.

Acknowledgements

This work is partially supported by the National Scientific Fund (Bulgaria) underGrant 815/98. B.R. would like to thank to J.-L. Leroy for the invitation and kindhospitality and to Z. Mouradian for kind permission to use the solar horizontaltelescope.


References

Balthasar, H., Stellmacher, G., and Wiehr, E.: 1988, Astron. Astrophys. 204, 286.Bashkirtsev, V. S., Kobanov, N. I., and Mashnich, G. P.: 1983, Solar Phys. 82, 443.Engvold, O.: 1972, Solar Phys. 23, 346.Engvold, O., Wiehr, E., and Wittman, A.: 1980, Astron. Astrophys. 85, 326.Giogolashvili, M. Sh. and Zhugzhda, Yu. D.: 1982, Solar Phys. 77, 95.Jensen, E.: 1982, Solar Phys. 77, 109.Jensen, E., Yi, Z., and Engvold, O.: 1994, Solar Phys. 149, 209.Joarder, P. S., Nakariakov, V. M., and Roberts B.; 1997, Solar Phys. 173, 81.Kraichnan, R. H.: 1965, Phys. Fluids 8, 1385.Liszka, L.: 1970, Solar Phys. 14, 354.Malville, J. M.: 1968, Solar Phys. 4, 313.Molowny-Horas, R, Oliver, R., Ballester, J. L., and Baudin, F.: 1997, Solar Phys. 172, 181.Mouradian, Z. and Leroy, J.-L.: 1977, Solar Phys. 51, 103.Newton, H. W.: 1934, Monthly Notices Roy. Astron. Soc. 94, 472.Oliver, R and Ballester, J. L.: 1996, Astrophys. J. 456, 393.Oliver, R., Ballester, J. L., Hood, A. W., and Priest, E. R.: 1993, Astrophys. J. 409, 809.Petrov, N. I., Dermendjiev, V. N., and Rompolt, B.: 1998, JOSO Ann. Report 1997, p. 145.Pettit, E.: 1932, Astrophys. J. 76, 9.Pojoga, S. and Molowny-Horas, R.: 1999, Solar Phys. 185, 113.Severny, A. B. and Khokhlova, V. L.: 1953, Izv. Krymsk. Astrofiz. Obs. 10, 9.Solov’ev, A. A.: 1985, Soln. Dannye No. 9, 65.Tandberg-Hanssen, E.: 1995, The Nature of Solar Prominences, Kluwer Academic Publishers,

Dordrecht, Holland, p. 95.Ten Bruggencate, P. and Elste, G.: 1958, Nature 182, 1154.Tomson, W. T. and Schmieder, B.: 1991, Astron. Astrophys. 243, 501.Wiehr, E., Stellmacher, G., and Balthasar, H.: 1984, Solar Phys. 94, 285.Yi, Z.: 1992, Doctoral Thesis, University of Oslo.Yi, Z. and Engvold, O.: 1991, Solar Phys. 134, 275.Yi, Z., Engvold, O., and Keil, S. L.: 1991, Solar Phys. 132, 63.Zirker, J. B., Engvold, O., and Yi, Z.: 1994, Solar Phys. 150, 81.

line-of-sight velocity fluctuations of a quiescent prominence

Documents