are activity characterized by 18 C-class and 2 M-class ares. Active region NOAA 8076 (BBSO 3877) was one of the rst active regions in the new Solar Cycle No.
HIGH-SPATIAL RESOLUTION OBSERVATIONS OF A SMALL -SPOT C. DENKER and H. WANG
Big Bear Solar Observatory, New Jersey Institute of Technology, 40386 North Shore Lane, Big Bear City, CA 92314, U.S.A.
Abstract. The Big Bear Solar Observatory (BBSO) has a long tradition of are observations. In this letter,
we would like to direct the reader's attention to observations of a small -spot which produced a moderate
are activity characterized by 18 C-class and 2 M-class ares. Active region NOAA 8076 (BBSO 3877) was one of the rst active regions in the new Solar Cycle No. 23. We present high spatial resolution white-light observations obtained on August 31st, 1997 with the speckle masking technique (Weigelt 1977; Weigelt & Wirnitzer1983; Lohmann et al. 1983). Almost di raction limited speckle reconstructions revealed the complex and highly dynamical behavior of a small emerging -con guration in the central part of NOAA 8076. We found strong shear ows and indications of strong transverse elds in the small -spot. The
are-producing mechanism for this small activity complex was very similar to that of the outstanding
are-producing region NOAA 5395 of March 1989 (Wang et al. 1991), however, on a completely opposite spatial scale. As a by-product, the speckle interferometric techniques provided information about the seeing quality at a site. We used von der Luhe's (1984) spectral ratio technique to estimate the Fried-parameter r0 . We measured a Fried-parameter of r0 = 9:0 0:7 cm and a maximum Fried-parameter of r0max = 10:3 cm indicating good seeing conditions during our observations. Key words: Sun: activity { Sun: ares { Sun: magnetic elds { techniques: speckle interferometry
1. Introduction This letter is related to two important topics in solar physics, (1) applying speckle interferometric technique to obtain di raction limited solar images and (2) understanding photospheric shear structure and motion in complicated -con gurations with di raction limited resolution. High resolution observations are extremely challenging for ground-based solar observations. Recent breakthroughs utilize the most advanced techniques like image selection (Scharmer 1989), speckle interferometry (Keller 1992; de Boer 1993; Denker 1997), and phase diversity techniques (Paxman et al. 1996). Recent seeing test at BBSO by the National Solar Observatory has demonstrated that BBSO could achieve di raction limited seeing conditions if the dome seeing problem is resolved (Beckers 1997). As a part of the test to evaluate the actual seeing condition at BBSO and to improve high spatial resolution observations, for the rst time in the BBSO's history, we recently used a fast (90 frames/s) 12-bit camera to carry out speckle observations. Near the end of August 1997, we used this technique to observe the properties of a small -sunspot which produces some levels of solar activity, including several C- and M-class
ares. Our motivation is to investigate shear structure and shear ow with di raction limited resolution. It has been found that magnetic structures are complicated in are productive active regions. With relatively high resolution, magnetic channels were found by Zirin & Wang (1993b), which are mixed in polarity and elongated in the direction of
2 C. DENKER AND H. WANG shear motion. The importance of magnetic shear to solar ares has been demonstrated by many authors (e.g. Tang & Wang 1993). Shear can be generated by two ways { the relative shear motion and head-on collisions of magnetic elements with opposite polarities (Wang 1992). The importance of -spot in producing solar ares were studied by Zirin & Liggett (1987). The shear ow has been studied for a number of big active regions, the most notable one is NOAA 5395 in March 1989 (Wang et al 1991). In this letter, we present our investigations of the shear structure in the center of a -con guration with di raction limited resolution, such an investigation has never been done.
2. Observations The observations were performed at the Big Bear Solar Observatory on August 30th and 31st, 1997. The optical setup for speckle interferometry was mounted on the center bench of the 65 cm refractor. To avoid saturation of the detector we used two neutral density lters (NG5 1 mm and NG9 1mm) to reduce the intensity by a factor of approximately 60. We observed white-light images at = 520 nm with a bandpass of = 6 nm. The spatial resolution of the refractor is, according to the Rayleigh-criterion = 1:22=D = 0:20100 , whereas the di raction limit is reached at =D = 0:16500 . A 256 by 254 12-bit Dalsa CCD camera is used as the detector. The data rate of the camera is approximately 90 frames/s which corresponds to an exposure time of about 10 ms. The camera is interfaced to a Pentium-166 Personal Computer containing a Tech imaging board for data acquisition and analyses. The computer has 128 MByte of RAM, so that up to 700 live images can be stored in the memory. The measured eld of view (FOV) was 25:000 24:800 which translates to an image scale of 0.098 arcsec/pixel. We started the observations at 20:58 UT. We took 24 sequences of 200 short exposure images. The time between successive frames was approximately 1 min. We operated the CCD camera in the master-mode to write the 200 frames as fast as possible to the RAM of the computer. After displaying the images, the frames were copied to hard disk. Additional sequences of 200 dark frames and
at eld frames were taken to compute an average dark and at eld frame. The at eld frames were obtained from an area with granulation close to the disk center while moving the telescope in the fast mode along both axis. For a proper noise calibration, we took 200 defocussed frames without any solar structures at disk center. Finally, the data were burned onto CD-ROM for further analysis at BBSO's UNIX workstation cluster. For a detailed morphological study, we also analyzed high-resolution Ca I (610.3 nm) ltergrams, H ltergrams, and Video-magnetograms. BBSO's synoptic full-disk Ca K II, H , and white-light observations were helpful to compare the morphology of the ne structures with the overall activity of the Sun.
HIGH-SPATIAL RESOLUTION OBSERVATIONS OF A SMALL -SPOT
3
11
r0 [cm]
10
9
8
7 1
5
10 15 Sequence No.
20
24
Figure 1. Temporal variation of the Fried-parameter r0 from 20:58 UT to 21:22.
3. Data Reduction The images were corrected by means of the average dark and at eld image. Due to imperfections in the at eld correction, some pixels still show strong deviations from their neighborhood. In these cases, the intensity value were replaced by the median of the adjacent pixels. We selected an area with granulation to normalize the images so that the mean value of the intensity is unity. We chose the image with the highest rms-contrast Cmax as a reference frame. We removed the image displacement with respect to that reference frame by means of a cross correlation analysis. Since the following speckle interferometric techniques are only valid for isoplanatic patches, we had to divide the specklegrams in an mosaic of 6 partially overlapping images of 64 pixels corresponding to an area of 6:300 6:300 . To derive the Fried-parameter r0 for a sequence, we applied the spectral ratio technique (von der Luhe 1984) to the 36 200 subframes. The temporal evolution of the Fried-parameter r0 for all the 24 sequences is shown in Figure 1. Once the Fried-parameter is known, we used Labeyrie's (1970) classical method to obtain the amplitudes of the object's Fourier transform and the speckle masking method (Weigelt 1977; Weigelt & Wirnitzer 1983; Lohmann et al. 1983) to derive the phases of the object's Fourier transform. A sensitive noise lter (de Boer 1996) was applied during the calculation of these phases. Back-transformation of the modulus and phases of the object's Fourier transform yields a mosaic of 6 6 partially overlapping images. Finally, we aligned these subframes and put them very accurately together.
4
C. DENKER AND H. WANG
a
N
b
E
Figure 2. (a) Quasi continuum image of NOAA 8076 observed in the line wing of the Ca-line at = 610:3 nm and (b) corresponding magnetogram at 21:01 UT on August 31st, 1997. The white boxes outline the speckle reconstruction and the enlarged magnetogram shown in Figure 3a and 3b. The eld of view amounts to 245 190 . The orientation of the following gures is the same as in this gure. 00
00
4. Results The acquisition of a very fast Dalsa CCD camera enabled us to implement speckle interferometric techniques for the rst time at BBSO. Figure 1 shows the variation of the Fried-parameter r0 , and demonstrates that the seeing was good during the observations and excellent in some cases (cf. von der Luhe 1984), especially at the beginning of the 25-minute time series. We measured an average Fried-parameter of r0 = 9:0 0:7 cm and a maximum Fried-parameter r0max = 10:3 cm. Examining power spectra and visual inspection of the speckle reconstructions disclose ne structures close to the di raction limit of the 65 cm vacuum re ector. The time series of speckle reconstructions presented in this letter represents one of the most highly resolved observations obtained at the Big Bear Solar Observatory, so far.
5
HIGH-SPATIAL RESOLUTION OBSERVATIONS OF A SMALL -SPOT
a
b Figure 3. (a) Speckle reconstruction of the -spot con guration observed at 20:59 UT on August 31st, 1997. (b) Magnetogram of the same region at 21:01 UT. The tickmarks are separated by 1 . 00
During the disk passage of NOAA 8076, 18 C-class and 2 M-class ares were observed. The region passed the central meridian in the northern hemisphere at a latitude of 30 on August 30th, 1997. The region has grown rapidly at the beginning and exhibited bright H plage. It showed an abnormal E-W neutral line with mixed polarity spots emerging in the intermediate portion of the sunspot group. The region was emerging at the location of an existing activity complex that produced several ares in early August. NOAA 8076 reached its maximum on August 30th, and developed a small -con guration with mixed polarity in the central part of the region on August 31st. Figure 2a shows a quasi continuum image of NOAA 8076 obtained with BBSO's video-magnetograph system in the line wing of the Ca I line at = 610:3 nm. The corresponding magnetogram is depicted in Figure 2b. The sunspot group consists of one major sunspot of preceding polarity, several smaller spots with rudimentary or complex penumbrae and various pores. The most conspicuous feature of this region is an area with spots and pores of mixed polarity within the trailing part of the region. The Space Environment Center (SEC) labeled NOAA 8076 according to the Mt. Wilson classi cation from August 29th to August 31st, and on September 2nd as a
6
C. DENKER AND H. WANG
F4
P3
F1
P4
P2 F3 P1
F5
F2
Figure 4. Schematic sketch of the small -spot con guration with indicators of the proper motions visible in the magnetograms and the high resolution white-light images.
-sunspot group { a bipolar group in which one spot was out of place relative to a typical
bipolar group. During the remainder of the time, NOAA 8076 was classi ed as a normal bipolar group. Inspecting the high spatial resolution speckle reconstructions as shown in Figure 3a, we nd several pores squeezed in length embedded in the complex penumbra surrounding spots F1 and F3 (cf. the schematic sketch of Figure 4 for the nomenclature used to mark out the sunspots and pores). Therefore, the activity center of NOAA 8076 may just as well be classi ed as a -spot { even though a very small sized -spot. The highly dynamical behavior in the immediate neighborhood of the activity center is clearly discernible in a time-lapse movie of the speckle reconstructions. Unfortunately, only limited disk space was available on the computer during the rst tests with the Dalsa CCD camera resulting in a relatively short time series of 24 minute duration. The most conspicuous dynamical behavior discernible in the cadence occurred between the two dominant sunspots F1 and F3 which exhibit an irregular and complex penumbrae with embedded pores. These elongated pores, of 3 to 5 arcsec length and a width of around 100 , provide one key to understand the are producing mechanism. Examining time lapse movies of magnetograms for August 31st, we nd that spot F3 moved slower than the rest of the active region resulting in an eastward motion relative to the group and thereupon pushing through an area of opposite polarity. This head-on collision (cf. Wang 1992) squeezes the p-polarity pores so that they become elongated in shape and the pores expeditiously move around sunspot F3 generating strong shear ows. In this process, the bright laments separating the p-polarity pores and sunspot F3 become as bright as 1.3
7
HIGH-SPATIAL RESOLUTION OBSERVATIONS OF A SMALL -SPOT 1.20 1.10
I / I0
1.00 0.90 0.80 0.70 0.60 0
1
2
3 y [arcsec]
4
5
6
Figure 5. Intensity cut through the dark horizontal channels shown in Figure 3a.
times the intensity of an area with granulation. A similar con guration has been described by Wang el al. (1991) but for one of the largest observed regions. In the upper third of Figure 3a, dark horizontal channels separate elongated bright features. The bright features have a length of about 200 and width of approximately 0.500 . The four horizontal arrows in the upper half of Figure 4 indicate the proper motion of the elongated granules along the dark horizontal channels towards sunspot F1 and pore F4 , respectively. Narrow lanes of transverse magnetic elds have been reported by Zirin & Wang (1993b). However, since we only deal with observations of longitudinal magnetic elds, some precautions against simply identifying these dark horizontal channels with lanes of transverse elds are appropriate. De ning pores as dark magnetic features lacking a penumbra may be appropriate in simple cases but in our case, where several magnetic features of opposite polarity co-exist in a relatively small area, the distinction between sunspots and pores becomes dicult and somewhat arbitrary. Figure 5 shows an intensity pro le through the conspicuous dark horizontal lanes shown in Figure 3a. The position of the intensity pro le is indicated by a white vertical bar in Figure 3a. The length of the intensity pro le amounts to approximately 600 . In the middle of the plot, we nd four well de ned intensity peaks with a FWHM of 0.3 to 0.4 00 . The small FWHM demonstrates again the almost di raction limited resolution we achieved in the present observations. The intensity of the bright elongated feature resembles bright penumbral laments, but we do not nd any comet like features or penumbral grains moving toward the spots as it would be typical in larger, more regular sunspots. The region with lower intensity on the very
8
C. DENKER AND H. WANG
Figure 6. H high resolution image observed at 22:22 UT. The eld of view amounts to 125 125 . The image was slightly -corrected to enhance the details in the background. 00
00
left side of Figure 5 refers to one of the elongated pores of p-polarity. Another remarkable feature attracts the attention in Figure 3a. Very faint, bright streamers spiral from the west into sunspot F3 as pointed out by the arrows with white arrowheads in Figure 4. The upper part of sunspot F3 is hidden beneath this thin silken veil. The low contrast of these features as well as the very complex motions in the time series are the reason why Local Correlation or Feature Tracking Routines failed or at least gave not very reliable results. Therefore, we choose display the proper motions in a schematic sketch (Figure 4) which were inferred by closely inspecting the 25-minute time-lapse movie of the speckle reconstructions. The time series of high resolution white-light images was enclosed by two C-class ares that recurred in nearly identical form. Ellison et al. (1960) have proposed for such events the term `homologous ares'. The rst are began at 20:26 UT and reached its maximum brightness at approximately 20:32 UT. Since we switched to the set-up for speckle interferometry at the 65 cm re ector, we had to interrupt the H observations, so that we were not able to observe the decaying phase of the are. The second C-class event took place from 22:15 UT to 22:36 UT with a maximum at 22:22 UT. Some short lived, ickering brightenings close to the location of the are's footpoints could be observed up to 15 minutes after the are decayed. Figure 6 shows an H center line image of the second C-class
are at 22:22 UT, approximately one hour after ending the speckle observations. In the H ltergram, three footpoints of the are are easily discernible. The central footpoint is located in the neighborhood of the dark horizontal channels and close to the pores P2 and P3 where we found strong shear ows. The two observed homologous C-class ares are
HIGH-SPATIAL RESOLUTION OBSERVATIONS OF A SMALL -SPOT
9
most likely X-type ares. The missing fourth footpoint might be very close to sunspot F3 where we nd strong, converging magnetic elds which prevent precipitation of electrons (Wang et al. 1995).
5. Conclusions Combining speckle interferometric techniques with magnetograms and H ltergrams provided a powerful tool to study complex magnetic eld con gurations and mechanisms that produce ares. However, magnetograms still lack sucient spatial resolution (cf. Figure 3a and 3b). Therefore, magnetograms with sub-arcsec resolution are highly desirable since we found evidence in the present investigation that changes in the magnetic eld con guration might be important even on these small scales. The rst successful steps in obtaining magnetograms with sub-arcsec resolution direction were performed with a new digital magnetograph system at BBSO (Wang et al. 1997). The core of this system is a CCD camera with a high frame rate which allows to overcome the disturbing e ects of the turbulent Earth's atmosphere by various post-processing techniques (frame selection, image alignment, sensitive noise lters, and speckle interferometry). The present results encourage further studies of activity centers in highly active and are producing regions, especially during the rise of the new Solar Cycle No. 23.
Acknowledgements We gratefully acknowledge the provision of the speckle code by Dr. C.R. de Boer. We thank Randy Fear, Bill Marquette, Je Nenow, and the sta of the Big Bear Solar Observatory for their support during the observations. This work was supported by AFOSR-95-0070, NAG-5-4919, NSF-ATM-96-28862, and NSF-ATM-97-96196.
References Beckers, J.: 1997, Bull. Am. Astron. Soc. 29/2, 897. de Boer, C.R.: 1993, Speckle-Interferometrie und ihre Anwendung auf die Sonnenbeobachtung, PhD-Thesis, Gottingen. de Boer, C.R.: 1996, Astron. Astrophys. Suppl. Ser. 120, 195. Denker, C.: 1997, Astron. Astrophys. 323, 599. Denker, C.: 1996, Untersuchung der Feinstruktur von Sonnen ecken mit Methoden der SpeckleInterferometrie, PhD-Thesis, Gottingen. Denker, C., de Boer, C.R., Volkmer, R., Kneer, F.: 1995b, Astron. Astrophys. 296, 567. Ellison, M.A., McKenna, S.M.P., Reid, J.H.: 1960, Dunsink Obs. Publ. 1, 1. Keller, C.U.: 1992, Nature 359, 307. Labeyrie, A.: 1970, Astron. Astrophys. 6, 85. Lohmann, A.W., Weigelt, G.P., Wirnitzer, B.: 1983, Appl. Opt. 22, 4028. Paxman, R.G., Seldin, J.H., Lofdahl, M.G., Scharmer, G.B., Keller, C.U.: 1996, Astrophys. J. 466, 1087. Scharmer, G.B.: 1989, in R.J. Rutten and G. Severino (eds.), Solar and Stellar Granulation, NATO ASI Ser. C 263, 161.
10
C. DENKER AND H. WANG
Tang, F., Wang, H.: 1993, Sol. Phys. 143, 107. von der Luhe, O.: 1984, J. Opt. Soc. Am. A 1, 510. Wang, H.: 1992, Sol. Phys. 140, 85. Wang, H., Denker, C., Spirock, T., Yang, S., Goode, P.: 1997, Bull. Am. Astron. Soc. 29/2, 919. Wang, H., Gary, D.E., Zirin, H., Schwartz, R.A., Sakao, T., Kosugi, T., Shibata, K.: 1995, Astrophys. J. 453, 505. Wang, H., Tang, F., Zirin, H., Ai, G.: 1991, Astrophys. J. 380, 382. Weigelt, G.: 1977, Opt. Comm. 21, 55. Weigelt, G., Wirnitzer, B.: 1983, Opt. Lett. 8, 389. Zirin, H., Marquette, W.: 1991, Sol. Phys. 131, 149. Zirin, H., Liggett, M.A.: 1987, Sol. Phys. 113, 267. Zirin, H., Wang, H.: 1993a, Sol. Phys. 144, 37. Zirin, H., Wang, H.: 1993b, Nature 363, 426.
