OBSERVATIONS OF SUNSPOT UMBRAL OSCILLATIONS T. HORN, J. STAUDE and V. LANDGRAF Astrophys. Institut Potsdam, Sonnenobservatorium Einsteinturm, Telegrafenberg, D-14473 Potsdam, Germany (Received 10 July 1996; accepted 24 February 1997) Abstract. The solar vacuum telescopes VTT and GCT at Tenerife have been used to obtain highresolution two-dimensional spectro-polarimetric observations of oscillations in the photospheric layers of sunspots. At the GCT the area of the sunspot has been scanned by shifting the spectrograph slit; at the VTT a Fabry Pérot interferometer has been applied to get narrow-band filtergrams directly and to scan through the line profile. The spectra of velocity oscillations show the known features of closely packed power peaks in bands of periods around 3 min (strengthened) and 5 min (weakened with respect to the quiet Sun). In the same frequency bands the more reliable VTT data show significant oscillations of the magnetic field strength as well, which could not be attributed to disturbing influences. Maximum power of both velocity and magnetic oscillations and a strong correlation between them, in the 3-min band in particular, is found to occur in those parts of the umbra where the magnetic lines of force are parallel to the line of sight. The oscillations are characterized by a marked spatial fine structure and a non-stationary behaviour. 1. Introduction A sunspot is a unique laboratory to investigate the dynamics of a magnetized atmosphere and of magneto-atmospheric waves in particular. Observations in spectral lines formed at different height levels of the umbral atmosphere show oscillations of velocity (Doppler shifts) and of intensity, that is, of thermodynamic quantities. The analyses of the time series often display sharp power peaks which are closely packed and concentrated in several period bands, without showing significant power between them. In umbrae there seem to exist two or three bands with periods around 3 min, 5 min, and possibly &20 min, the oscillations of which are likely produced by different physical mechanisms. A comprehensive review on observations of sunspot oscillations has been given by Lites (1992); possible approaches to an interpretation have been summarized, e.g., by Zhugzhda, Locans, and Staude (1987), Thomas and Weiss (1992), and Staude (1994). The results of attempts to measure oscillations of the umbral magnetic field are contradictory. There are a few papers reporting oscillatory power of single magnetic field components in the 3-min and 5-min period bands, often without a clear correlation with other oscillatory phenomena (Mogilevskij, Obridko, and Shelting, 1972, 1973; Gurman and House, 1971; Efremov and Parfinenko, 1996), but most observers were unable to discover any significant field oscillation (Schultz and White, 1974; Bachmann, 1983; Thomas, Cram, and Nye, 1984). Two-dimensional data with high spatial resolution are rare. Diagnostic, k h diagrams ( = P,1 is the frequency of the oscillation, P its period, and k h its Solar Physics 172: 69 76, 1997. c 1997 Kluwer Academic Publishers. Printed in Belgium.
70 T. HORN, J. STAUDE, AND V. LANDGRAF horizontal wave number) of velocity fluctuations have been obtained only twice: Abdelatif, Lites, and Thomas (1986) observed significant power only for small and k h but strong attenuation to the right-hand side of a line corresponding to a phase velocity of 25 km s,1 which is close to the Alfvén speed c A in the umbral photosphere. Penn and LaBonte (1993) found consistency with the ridge structure of the global p-mode power outside the spot. This is why we tried to get new data of sunspot oscillations by means of the vacuum solar telescopes at the Observatorio del Teide, Tenerife, using different techniques of scanning the image or the spectral line. 2. Observations and Analysis GCT. The Gregory coudé telescope (GCT) at Tenerife has been applied to observe the umbra of the main f-spot of the active region NOAA 6895 (23 N, 40 W) on September 3, 1991; the time series covers the time 09:40 10:25 UT corresponding to good seeing conditions with an image blurring of.1 00.Stokes (I V ) spectra have been obtained in two spectral regions around 5250 Å and in the Na D 2 line. That is, the light beam passed the polarimeter and a spectrum cutter, forming 4 partial spectra on one CCD matrix with 1024 1024 pixels, to which a binning of 2 2 has been applied. The present study is restricted to the line Fe I 5250.2 Å, that is to the upper photosphere. The spectral resolution was =6mÅ, the spectral slit width was 60 m ˆ= 10mÅ; the spatial sampling interval along the slit was x = 0:33 00, resulting in 1 00 after averaging over 3 pixels in the final data reduction. The sunspot has been scanned by shifting the spectrograph slit perpendicular to its length in 3 steps of 2 00 each; the 4 slit positions were controlled by slit-jaw pictures, while the solar rotation has been compensated for. Exposures have been obtained every 15 s. Problems related to the image rotation by the Gregory coudé system and some inaccuracy of the telescope guiding could only be overcome by interpolating in the image. The derived data are time series and power spectra of the fluctuations of velocity v(t) from Doppler shifts, of the magnetic field B(t)from Zeeman splittings (the splittings have been measured by fitting the line core by a parabola and measuring the distance of the components in the I V images), and of the continuum intensity in the umbra Ic;u(t) as well as in the quiet Sun (outside the sunspot) Ic;q(t) for comparison. VTT. The Vacuum Tower Telescope (VTT) at Tenerife has been used together with a polarimeter in front of a two-dimensional imaging spectrometer with a Fabry Pérot interferometer (FPI). The spectrometer has been described by Bendlin, Vokmer, and Kneer (1992) and Bendlin and Volkmer (1995), the remaining instrumentation, calibration, and data reduction by Horn, Hofmann, and Balthasar
UMBRAL OSCILLATIONS 71 Figure 1. GCT data. Magnetic field strength B(t) (full line) and umbral continuum intensity Ic;u(t) (dashed line) measured in an umbral position. The umbral contrast is nearly proportional to Ic;u with = 0:2 atic;u 3:0 10 5 counts. The similarity between the measured variations of both B (t) and (t) means that B (t) is mainly determined by variations of parasitic stray light. (1996). The narrow-band filtergrams were recorded on a 286 384 pixel CCD camera with 0:2 00 pixels. Image blurring resulted in a real spatial resolution of better than 0:8 00 for the best images. The polarimeter is able to measure the full Stokes vector, however, in the present work we concentrated on the circularly polarized components (I V ). The spectral scanning was accomplished by tuning the FPI with a sampling rate of = 10:9mÅ, covering the whole line profile in 54 s. The field of view covers 55 00 37 00. Although the s=n ratio of the spectra is only 1% in the intensity, the setup is able to measure line shifts with good accuracy (e.g., Balthasar et al., 1996). The present observations were obtained on July 20, 1994, and were focused on the main spot of the active region NOAA 7757, 30 NE from the disk center. The photospheric spectral line Fe I 6173.4 Å has been used. The time series covers 127 scans of the spot within 114 min. White-light images from the same field of view were taken strictly simultaneously to all narrow-band pictures. The derived data cover pairs of Stokes (I V ) profiles for each pixel in the field of view and for each moment, from which v(t) and B(t) (the measuring procedure is similar to that for the GCT data) and the corresponding maps of oscillatory power have been derived, moreover, the Ic(t) maps are available for comparison and calibration purposes. 3. Discussion and Conclusions GCT. The resulting power spectra of v at different umbral positions are similar to those obtained with the VTT described below; they will not be discussed here in detail. Reliable oscillations of B, however, have not been discovered in the GCT
72 T. HORN, J. STAUDE, AND V. LANDGRAF Figure 2. VTT data. Spatial distribution of power of the oscillations of B (t) (left),v (t) (middle) and phase difference (right) for two selected bands; the upper row shows the direction towards the limb, the power gray scale (maximum power is brightest), and an Ic picture of the same region. The gray coding for the map is: black for a coherence <0.85, elsewhere gray for =2and white for 0.
UMBRAL OSCILLATIONS 73 Figure 3. VTT data. Power spectra and the related time series of v (t) for selected positions in the umbra and the adjacent photosphere for comparison; the positions are marked and numbered in the Ic picture of Figure 2. The dashed horizontal lines are the 99 % confidence limits corresponding to 2:33 standard deviation according to Groth (1975).
74 T. HORN, J. STAUDE, AND V. LANDGRAF Figure 4. VTT data. Power spectra and the related time series of B (t) for selected positions in the umbra and the adjacent photosphere for comparison; the positions are marked and numbered in the Ic picture of Figure 2. The dashed horizontal lines are the 99 % confidence limits corresponding to 2:33 standard deviation according to Groth (1975).
UMBRAL OSCILLATIONS 75 data: apparent oscillations were not correlated with those of v but rather produced by variations of the umbral intensity contrast = Ic;u=Ic;q due to fluctuations of parasitic stray light which were quite large in spite of the good seeing conditions. For example, see the time series in a selected umbral position shown in Figure 1. VTT. From the (I + V ) and (I, V ) images the Doppler shifts and the Zeeman splittings of the observed line were derived by fitting a polynomial to the line centres of the -components. The measured splittings are influenced by the total field strength B and to a minor extent by the angle of inclination between the line of sight and the vector B, if the splitting is incomplete. Figure 2 shows maps of power of v and B in two -bands (around P 3 min and 5 min, respectively), moreover, the coherence and phase difference between both types of oscillations are given. Figures 3 and 4 display power spectra and time series of v(t) and B(t), respectively, at selected positions which are marked and numbered in the sunspot Ic picture in Figure 2. The power spectra of v(t) in the umbra show the known features of strong power in bands of period around 3 min (strengthened) and 5 min (weakened with respect to the quiet Sun). Contrary to the GCT data the more reliable VTT data show significant power of oscillations of B(t) as well, which is concentrated in the same frequency bands as the power of v(t). To exclude a possible influence of the jitters of the telescope the intensity fluctuation of a small area at the umbra penumbra boundary has been checked: significant power is observed at frequencies <1:5 mhz only. An inspection of the observed umbral contrast shows a large amount of stray light, similar to that in the GCT, but contrary to the latter case it is very stable in time: in the VTT data the power density of the fluctuations of is very weak and equally distributed over all. The spatial distribution of the power of v(t) and B(t) is inhomogeneousacross the umbra and shows marked spatial fine structures. Maximum power is measured in those parts of the umbra (close to the umbra penumbra boundary) where we are looking along the lines of force of B, thus demonstrating the longitudinal character of the oscillations with respect to the direction of B. In the same parts of the umbra we found also the largest correlation and 0 between both types of oscillations, in the 3-min band in particular. The penumbral regions show rather an anti-correlation in the 5-min band. The time series clearly demonstrate the non-stationary character of the oscillations. In the further analysis of the data we plan to investigate in more detail the significance of the various resonant power peaks and the non-stationary behaviour of the oscillations. Such information will provide a better base for checking theoretical predictions. At the GCT the reliability of the data will be improved by introducing a polarization-free image derotator and a polarization-free scanning system which will no longer require to move the heavy telescope (Koschinsky and Kneer, 1996).
76 T. HORN, J. STAUDE, AND V. LANDGRAF Acknowledgements The authors gratefully acknowledge support of the present work by the Deutsche Forschungsgemeinschaft (DFG) under grants Sta 351/2-1, 2-2, 3-1, and 3-3. We are grateful to Cornelia Bendlin and to Franz Kneer for advice and help with the observing techniques at Tenerife. The GCT and FPI are operated by the Universitäts-Sternwarte, Göttingen, the VTT by the Kiepenheuer-Institut für Sonnenphysik, Freiburg, at the Spanish Observatorio del Teide of the Instituto de Astrofísica de Canarias. References Abdelatif, T. E., Lites, B. W., and Thomas, J. H.: 1986, Astrophys. J. 311, 1015. Bachmann, G.: 1983, Phys. Solariterr. Potsdam 20, 29. Balthasar, H., Schleicher, H., Bendlin, C., and Volkmer, R.: 1996, Astron. Astrophys. 315, 603. Bendlin, C. and Volkmer, R.: 1995, Astron. Astrophys. 112, 371. Bendlin, C., Volkmer, R., and Kneer, F.: 1992, Astron. Astrophys. 257, 817. Groth, R. C.: 1975, Astrophys. J. Suppl. 29, 285. Gurman, J. B. and House, L. L.: 1981, Solar Phys. 71, 5. Horn, T., Hofmann, A., and Balthasar, H.: 1996, Solar Phys. 164, 321. Koschinsky, M. and Kneer, F.: 1996, Astron. Astrophys. Suppl. Ser. 119, 171. Landgraf, V.: 1995, Diploma Thesis, Berlin. Lites, B. W.: 1992, in J. H. Thomas and N. O. Weiss (eds.), Sunspots: Theory and Observations, NATO ASI Series C375, 261. Mogilevskij, E. I., Obridko, V. N., and Shelting, B. D.: 1972, Astron. Tsirk. 669, 1. Mogilevskij, E. I., Obridko, V. N., and Shelting, B. D.: 1973, Radiofizika 16, 1357. Penn, M. J. and LaBonte, B. J.: 1993, Astrophys. J. 415, 383. Schultz, R. B., and White, O. R.: 1974, Solar Phys. 35, 309. Staude, J.: 1994, in R. J. Rutten and C. J. Schrijver (eds.), Solar Surface Magnetism, NATO ASI Series C433, 189. Thomas, J. H. and Weiss, N. O. (eds.): 1992, Sunspots: Theory and Observations, NATO ASI Series C375, 3. Thomas, J. H., Cram, L. E., and Nye, A. H.: 1984, Astrophys. J. 285, 368. Zhugzhda, Y. D., Locans, V., and Staude, J.: 1987, Astron. Nachr. 308, 257.