1: \documentclass[apj]{emulateapj}
2:
3: \newcommand{\FeI}{\ion{Fe}{1}}
4: \newcommand{\CaII}{\ion{Ca}{2}}
5: \newcommand{\kms}{km$\,$s$^{-1}$}
6:
7:
8:
9: \newcommand{\aliemail}{ali@nso.edu}
10: \newcommand{\naemail}{na.deng@csun.edu}
11: \newcommand{\debriemail}{debiprasad.choudhary@csun.edu}
12: \newcommand{\haiminemail}{haimin@flare.njit.edu}
13: \newcommand{\chungemail}{cliu@bbso.njit.edu}
14: \newcommand{\cdenkeremail}{cdenker@adm.njit.edu}
15: \newcommand{\carstenemail}{cdenker@aip.de}
16:
17: \shorttitle{Flow Field Evolution of a Decaying Sunspot}
18: \shortauthors{Deng et al.}
19:
20:
21: \begin{document}
22:
23:
24: \title{Flow Field Evolution of a Decaying Sunspot}
25:
26: \author{Na Deng and Debi Prasad Choudhary}
27: \affil{California State University Northridge, Physics
28: and Astronomy Department, 18111 Nordhoff St., Northridge, CA 91330; \naemail, \debriemail}
29:
30: \author{Alexandra Tritschler}
31: \affil{National Solar Observatory/Sacramento Peak\footnote{Operated by the Association of Universities for Research in Astronomy, Inc. (AURA), for the National Science Foundation},
32: P.O.~Box 62, Sunspot, NM 88349, U.S.A.; \aliemail}
33:
34: \author{Carsten Denker\altaffilmark{1}}
35: \affil{ Astrophysikalisches Institut Potsdam, An der Sternwarte 16,
36: D-14482 Potsdam, Germany; \carstenemail}
37:
38: \author{Chang Liu and Haimin Wang}
39: \affil{New Jersey Institute of Technology, Physics Department,
40: Center for Solar-Terrestrial Research, 323 Martin Luther King Blvd, Newark,
41: NJ 07102 \\
42: Big Bear Solar Observatory, 40386 North Shore Lane, Big Bear
43: City, CA 92314, U.S.A.; \chungemail, \haiminemail}
44:
45: \altaffiltext{1}{New Jersey Institute of Technology, Physics Department,
46: Center for Solar-Terrestrial Research, 323 Martin Luther King Blvd, Newark,
47: NJ 07102, U.S.A.}
48:
49:
50: \begin{abstract}
51:
52: We study the evolution of the flows and horizontal proper motions in and around a decaying follower sunspot based
53: on time sequences of two-dimensional spectroscopic observations in the visible
54: and white light imaging data obtained over six days from June~7 to~12, 2005.
55: During this time period the sunspot decayed gradually to a pore.
56: The spectroscopic observations were obtained with the Fabry-P\'{e}rot based Visible-Light Imaging
57: Magnetograph (VIM) in conjunction with the high-order adaptive optics (AO) system operated
58: at the 65\,cm vacuum reflector of the Big Bear Solar Observatory (BBSO).
59: We apply local correlation tracking (LCT) to the speckle reconstructed
60: time sequences of white-light images around 600\,nm to infer horizontal proper motions
61: while the Doppler shifts of the scanned \FeI\ line at 630.15\,nm are used
62: to calculate line-of-sight (LOS) velocities with sub-arcsecond resolution.
63: We find that the dividing line between radial inward and outward proper motions in the
64: inner and outer penumbra, respectively, survives the decay phase.
65: In particular the moat flow is still detectable after the penumbra disappeared.
66: Based on our observations three major processes removed flux from the sunspot:
67: (a) fragmentation of the umbra, (b) flux cancelation of moving magnetic features (MMFs; of the same polarity as the sunspot)
68: that encounter the leading opposite polarity network and plages areas,
69: and (c) flux transport by MMFs (of the same polarity as the sunspot) to the surrounding network and plage regions
70: that have the same polarity as the sunspot.
71: \end{abstract}
72:
73: \keywords{
74: Sun: activity ---
75: Sun: photosphere ---
76: sunspots ---
77: instrumentation: high angular resolution ---
78: instrumentation: spectrographs ---
79: techniques: spectroscopic}
80:
81:
82: \section{Introduction}\label{sec:introduction}
83:
84: The evolution of active regions has been extensively studied for
85: decades \citep[e.g. ][and references therein]{solanki2003}. While
86: much progress has been achieved in understanding of how flux is
87: transported from deep in the convection zone to emerge at the solar
88: surface in form of sunspots, a comprehensive picture of the
89: regularities underlying the decay process
90: \citep[e.g.][]{martinezpillet2002} is still missing. Typically, a
91: statistical approach is followed to characterize the decay phase by
92: changes in umbral and penumbral brightness, magnetic flux, or area in
93: an attempt to find either a common (mean) decay law
94: \citep[e.g.][]{morenoinsertis+vazquez1988, petrovay+martinezpillet+vandrielgesztelyi1999}
95: or by studying the distribution of the decay rates
96: \citep[e.g.][]{martinezpillet+morenoinsertis+vazquez1993}. However, this strategy does only
97: evaluate the behavior of groups rather than of individual sunspots
98: and does not give detailed insight into the individual processes that
99: might contribute to or trigger the decay process.
100:
101: The decay phase of a sunspot can be initiated at any time
102: \citep{mcintosh1981} and proceeds differently for individual
103: sunspots. \citet{chapman+etal2003} analyzed the decay phase of many
104: sunspots and found a strong correlation between the sunspot
105: photometric decay rate, the total sunspot area and the umbral to total
106: area ratio. By carefully measuring the magnetic flux,
107: \citet{zwaan1992} detected no significant increase in the total
108: magnetic flux of the sunspot after the penumbra is
109: formed. Furthermore, \citet{leka+skumanich1998} observed that the
110: umbral flux does not decrease during the formation of a
111: penumbra. These findings suggest that the formation and decay of
112: penumbrae is not immediately related to changes in the sunspot
113: magnetic flux, but rather a phenomenon related to a change in the
114: local flow field in and around the flux concentration. Differences
115: between the flow field around the sunspot and that around
116: the umbral fragments and pores thus might allow to determine critical
117: conditions necessary for both penumbral formation and decay.
118:
119: The most prominent flow spectroscopically measured in a sunspot is
120: the Evershed flow \citep{evershed1909}, which reveals itself in form
121: of shifts and asymmetries of solar spectral lines and which is
122: interpreted as an outward mass flow in the photospheric layers of
123: penumbrae. The Evershed effect is intrinsically coupled to the
124: presence of a penumbra and ceases at the immediate outer penumbral
125: border. \citet{leka+skumanich1998} and \citet{yang+etal2003} detected
126: an Evershed flow as soon as the magnetic field lines reached a critical
127: inclination and organized filamentary structures became visible. The
128: Evershed flow itself varies quasi-periodically in time
129: \citep{shine+etal1994, rimmele1994, rouppe2003,
130: cabrera+etal2006a}, which is predominately caused by velocity
131: patches that propagate along penumbral filaments from the middle penumbra to the
132: outer penumbra, the so-called Evershed clouds (ECs) which were first observed and
133: named by \citet{shine+etal1994}. In their most recent study,
134: \citet{cabrera+etal2007} identify two classes of ECs, those that
135: vanish directly at the outer penumbral boundary and those that cross
136: the visible border and enter the moat flow. Once detached from the
137: penumbra, these ECs survive $\sim$14\,min before they disappear in
138: close proximity to the spot ($\sim$2\arcsec). The outward migration
139: of the ECs could give a natural explanation for the existence of a dividing
140: line within the penumbra separating features that move inwards
141: in the inner penumbra and outwards in the outer penumbra. This
142: dividing line appears typically in horizontal proper motion maps derived from either local
143: correlation tracking \citep[e.g.][]{november1986, november+etal1986}
144: or pattern recognition \citep[e.g.][]{sobotka+brandt+simon1999,
145: bovelet+wiehr2003}. \citet{molownyhoras1994} was the first to
146: identify such a ring-like dividing area, featuring near-zero
147: horizontal speed and constant velocity divergence.
148:
149: The sunspot moat is a zone that surrounds most mature sunspots and
150: some large sunspot pores. The moat region is characterized by a
151: radially directed outward mass flow. Although the moat is free of
152: stationary magnetic fields, time sequences of magnetograms show moving
153: magnetic features (MMFs), i.e., small-scale magnetic flux concentrations of
154: mixed polarity, that move along with the moat flow in radial
155: direction away from the sunspot \citep[]{sheeley1969,
156: harvey+harvey1973, brickhouse+labonte1988, yurchyshyn+wang2001, hagenaar+shine2005b,
157: kubo+shimizu+tsuneta2007, choudhary+bala2007}. There is increasing evidence that the
158: Evershed flow, the moat flow, and MMFs are closely related to each other
159: \citep[e.g.][]{sainzdalda+martinezpillet2005, ravindra2006,
160: cabrera+etal2006a}. In a case study
161: \cite{vargasdominguez+etal2007} find that for a complex active region
162: the existence of the moat flow is coupled to the presence of
163: penumbral filaments. Hence, the moat should neither
164: be found around individual pores or umbral fragments after
165: the penumbra disappeared. However, observational results differ considerably
166: depending on what type of sunspot and pores have been observed
167: and what techniques have been used
168: \citep[see e.g.][]{wang+zirin1992, denker1998, yang+etal2003}.
169:
170: In this paper, we study the evolution of line-of-sight (LOS) flows and horizontal proper motions in and around a
171: decaying sunspot. Our investigation is based on sequences of
172: high-spatial resolution observations obtained over six days with new
173: post-focus instrumentation at the Big Bear Solar Observatory (BBSO)
174: \citep{denker+etal2005, denker+tritschler2005, denker+etal2007a}. Both
175: the LOS velocities measured from two-dimensional spectroscopic
176: observations and horizontal proper motions derived with local correlation tracking (LCT) technique based on speckle reconstructed images are analyzed. In \S\ref{sec:observation}
177: we overview the two-dimensional spectroscopic and speckle imaging
178: observations. The data reduction is described in \S\ref{sec:reduction}
179: and the results are presented in \S\ref{sec:results}.
180: We discuss the results and offer our conclusions in \S\ref{sec:discussion}.
181:
182:
183: \section{Observations}\label{sec:observation}
184:
185:
186: \begin{figure}[t]
187: \epsscale{1.}
188: \plotone{f1.eps}
189: \caption{\textsl{a}: Speckle reconstructed white-light image of the follower sunspot in
190: active region NOAA\,10773 on 2005 June\,7. \textsl{b}: Corresponding continuum
191: intensity image observed with the two-dimensional spectrometer. The white arrow points towards
192: disk center (DC). \textsl{c}: Sample line profiles of the two-dimensional spectroscopy
193: averaged over small areas in the center-side (CS) and limb-side (LS)
194: penumbra, which are indicated by white boxes in \textsl{b}.}
195: \label{FIG3.1}
196: \end{figure}
197:
198:
199: The observations were performed with BBSO's 65\,cm vacuum reflector
200: in conjunction with the high-order adaptive optics (AO) system \citep{rimmele+etal2004,
201: denker+etal2007a}. We observed the isolated follower spot of active
202: region NOAA\,10773 during its disk passage
203: on five days in summer 2005: June\,7 ($\mu \approx
204: 0.96$), June\,8 ($\mu \approx 0.95$), June\,10 ($\mu \approx 0.77$), June\,11
205: ($\mu \approx 0.61$), and June\,12 ($\mu \approx 0.43$). Two-dimensional
206: spectroscopic scans of the photospheric spectral line \FeI\ 630.15\,nm
207: were obtained with the Visible-light Imaging Magnetograph (VIM) system,
208: in which a single Fabry-P\'erot etalon is the core element \citep{denker+tritschler2005}.
209: VIM has a bandpass of 8\,pm. A 0.3\,nm wide prefilter centered on the \FeI\ 630.15\,nm line
210: was used before the Fabry-P\'erot etalon. The \FeI\ line was sampled
211: at 101 equidistant wavelength points with a step size of 1.2\,pm corresponding
212: to a scanned wavelength interval of 0.12\,nm. The exposure time at each wavelength
213: point was 80\,ms. The cadence of each scan was about 18\,s. A
214: total of 100 scans can be obtained in 30\,min, which was the typical time range
215: of an observing run. The spectroscopic data have a field-of-view (FOV) of about $74\arcsec
216: \times 74\arcsec$. The 1k$\times$1k CCD camera was operated in a $2 \times
217: 2$\,pixel binning mode. Thus, the image scale was about 0.14\arcsec\,pixel$^{-1}$. The spatial resolution achieved by the spectroscopic scans is estimated at about 0.5\arcsec\ with power spectrum analysis.
218:
219: The spectroscopic observations are supported by quasi-synchronous bursts
220: of short-exposure (10\,ms) white-light images at 600\,$\pm$\,5\,nm
221: recorded with a fast CCD camera system for image restoration. Every 30\,s, a total of
222: 200 frames was captured at 15 frames\,s$^{-1}$, from which the
223: best 100 frames were automatically selected and combined to create one restored
224: image using a speckle masking algorithm \citep{denker+etal2005}. The
225: cadence of the reconstructed white-light images is thus 30\,s. A total of
226: 60 reconstructed images were obtained in a 30\,min observing run. The FOV is
227: about $70\arcsec \times 70\arcsec$ and the image scale is about
228: 0\farcs07\,pixel\,$^{-1}$. The spatial resolution of the reconstructed images is diffraction limited ( $\approx$ 0.25\arcsec) by the 65~cm reflector of the BBSO if we use $1.22\lambda/D$ as the measure of the diffraction limit. Each observing run typically lasted 30\,min.
229: Further details of the observations are given by
230: \citet{denker+etal2007a,denker+etal2007b}, who provide a
231: thorough description of the high-resolution observations and a performance
232: evaluation of the AO system.
233:
234: Figure~\ref{FIG3.1} shows the sunspot observed on June\,7 in a speckle
235: reconstructed white-light image (panel~\textsl{a}) and the corresponding continuum
236: intensity image obtained with VIM (panel~\textsl{b}). Panel~\textsl{c} shows two
237: sample line profiles averaged over a small area of
238: limb-side (LS) and center-side (CS) penumbra, which are indicated in panel~\textsl{b} by
239: white boxes. The two profiles are shifted w.r.t. each other
240: indicating the Evershed flow. The reconstructed image clearly shows the sunspot fine structure,
241: such as numerous umbral dots, the penumbral dark cores, and the
242: twist and writhe of the penumbral filaments. We also note that the penumbral
243: grains are brighter on the limb-side of the penumbra.
244:
245: \begin{figure}[t]
246: \epsscale{1.0}
247: \plotone{f2.eps}
248: \caption{Evolution of active
249: region NOAA\,10773 from 2005 June\,7 to June\,12. The
250: columns (from \textsl{left} to \textsl{right})
251: represent white-light (WL) images, LOS magnetic
252: field $B_{LOS}$, and \CaII\,K and H$\alpha$ filtergrams.}
253: \label{FIG3.2}
254: \end{figure}
255:
256:
257: NOAA\,10773 was also monitored by BBSO's 25\,cm refractor. These context observations include
258: vector magnetograms as well as H$\alpha$ and \CaII\,K filtergrams.
259: Figure~\ref{FIG3.2} shows the whole active region
260: and its evolution over six consecutive days based on
261: white-light information, LOS magnetograms, and \CaII\,K and H$\alpha$
262: filtergrams. BBSO's LOS magnetograms are produced by dividing the difference between the right- and left-hand circularly polarized Stokes components by their sum under weak field approximation. They suffer from the problem of Zeeman saturation when the field is stronger than typically 3000~G \citep{wang+etal1998,spirock2005}. We employ data from other observatories on some days because no synoptic BBSO data were available. For example, the white-light and magnetogram data on June\,9, 11, and 12 are from the
263: Michelson Doppler Imager \citep[MDI,][]{scherrer+etal1995} on board the Solar and
264: Heliospheric Observatory. The \CaII\,K and H$\alpha$ filtergrams on June\,11 and 12 are from the Mauna Loa Solar Observatory operated by the High Altitude Observatory. The active region was in the southern hemisphere (S15\arcdeg). The target was a relatively stable, round follower sunspot with positive magnetic
265: polarity located on the left side in the panels of Figure~\ref{FIG3.2}.
266: The leading spot on the right side was otherwise smaller in size. This configuration is an exception, since
267: the leading spot is typically more stable and larger in size. The bright
268: areas in \CaII\,K and H$\alpha$ filtergrams trace the plage regions that
269: are spatially correlated with magnetic fields of the supergranular network.
270:
271:
272: \section{Data Reduction}\label{sec:reduction}
273:
274: We apply standard procedures such as dark subtraction and gain table correction
275: to the spectroscopic observations and white-light images. The speckle
276: reconstructions and the scans are carefully co-aligned with each other.
277: The white-light images were also co-aligned with the corresponding MDI
278: intensity images in order to acquire accurate locations and orientations
279: of the sunspot on different days. In order to remove the 5-minute oscillation
280: and residual seeing distortions a subsonic filter with a phase velocity
281: threshold of 4\,km\,s$^{-1}$ was applied to the time-series of white-light
282: speckle and spectroscopic data.
283:
284: The LOS velocity is determined by a Fourier phase method
285: \citep[see][]{schmidt+stix+woehl1999}, which
286: uses the entire \FeI\, 630.15\,nm line profile and is very insensitive to noise. The telluric blend at 630.2~nm is not included in the calculation.
287: The determined shifts are converted to velocities by using
288: the Doppler formula. As a frame of reference we use the average
289: velocity of a small area in the darkest part of the umbra or
290: pore. The final calibrated dopplergrams follow the convention that redshifts are positive and blueshifts are negative. Thus, bright areas in the dopplergrams move away from the
291: observer, while dark areas move towards the observer. The \FeI\, 630.15\,nm
292: has a non-negligible Land\'e factor and is thus susceptible to
293: the presence of magnetic fields. The influence of the
294: magnetic field on the line shape, however, is symmetric with
295: respect to the line core and should not significantly influence
296: the velocity determination.
297:
298: \begin{figure}[t]
299: \epsscale{1.0}
300: \plotone{f3.eps}
301: \caption{Combination of LOS dopplergram and LCT horizontal proper motions of the observed sunspot
302: (NOAA~10773) on 2005 June\,7. Arrows show a 30-minute average of
303: LCT proper motion maps derived from
304: speckle-restored images. The gray-scale background image represents the
305: LOS velocity, which is the corresponding 30-minute
306: average of the dopplergrams. Redshifts are positive and correspond to
307: bright areas. The black contours outline the umbral and penumbral boundaries.}
308: \label{FIG3.3}
309: \end{figure}
310:
311:
312: \begin{figure*}[th]
313: \epsscale{1.1}
314: \plotone{f4.eps}
315: \caption{Sunspot evolution over six days during its decaying phase.
316: White contour lines mark the umbral and penumbral boundaries
317: determined from 60\,\% and 90\,\% of the quiet
318: Sun intensity, respectively. The yellow corks trace the dividing
319: lines between the inward
320: and the outward proper motions in the penumbra and around the residual pore. The red arrows in panel \textsl{f} point to the broken and distorted dividing line around the pore.}
321: \label{FIG3.4}
322: \end{figure*}
323:
324:
325: The ANA implementation (\url{http://ana.lmsal.com/ana/}) of the LCT algorithm \citep{november1986, november+etal1986} was applied to each of the individual 30-minute observing sequences of speckle
326: reconstructions. A Gaussian shape apodization window with a FWHM of 1.4\arcsec (corresponding to the mean size of the granulation) was used in the LCT technique. Since each sequence consists of 60 reconstructions, this resulted
327: in 59 individual LCT proper motion maps. The individual proper motion maps are averaged to
328: achieve a significant reduction of seeing induced random noise.
329: Furthermore, we use inverse cork maps to visualize sources of divergence in a LCT proper motion
330: map \citep[see e.g. ][]{molownyhoras1994, denker1998}. A total of 22,500 corks
331: (artificial tracer particles) were distributed evenly
332: across the average LCT proper motion map. Their movement is traced for 10\,h
333: backwards in time. The corks concentrate at positions where the proper motion shows strong and persistent divergence.
334: The resulting cork map traces the location of the dividing line between
335: inward and outward horizontal proper motions in the sunspot penumbra.
336: It should be mentioned that all the quantities and figures presented in this paper, unless otherwise noted, have been corrected for perspective foreshortening.
337:
338: To demonstrate the outcome of the described methodology Figure \ref{FIG3.3}
339: displays the average LCT proper motion map superposed on the LOS dopplergram
340: averaged over the same time period for 2005 June\,7.
341: It clearly shows the Evershed effect in the
342: background LOS dopplergram, and visualizes the inward motion in the inner penumbra
343: and outward motion in the outer penumbra as indicated by the arrows showing the horizontal proper motion.
344: The outward proper motion exceeds the outer penumbral boundary by 3\arcsec\ to
345: 5\arcsec\, while the LOS Evershed flow ceases more or less abruptly at the visible
346: boundary of the outer penumbra. Moreover, we detect diverging and converging trends
347: in the radial outward proper motions outside the sunspot.
348:
349:
350: \section{Results}\label{sec:results}
351:
352: Figure~\ref{FIG3.4} illustrates the result of the cork analysis
353: based on the reconstructed white-light images for all observing days.
354: The white contours mark the umbral and penumbral boundaries as
355: derived from intensity thresholds set to 60\,\% and 90\,\% of the quiet
356: Sun intensity, respectively. The yellow dots indicate the corks that are
357: swept to regions where the horizontal proper motion is not divergence free.
358: Figure~\ref{FIG3.4} clearly reveals the dividing lines, which separate the inward
359: and outward proper motions in the penumbra. The dividing lines always reside inside
360: the penumbra. They tend to be more symmetric when the sunspot penumbra itself has a more
361: symmetric shape (see panel \textsl{b} in Figure~\ref{FIG3.4}). They lose
362: symmetry when the sunspot penumbra becomes less symmetric (see the other panels
363: in Figure~\ref{FIG3.4}). Interestingly, even after the penumbra completely
364: disappeared, there is still a distinct dividing
365: line around the remaining pore, although, incomplete
366: and disrupted at some locations (see panel \textsl{f} in Figure~\ref{FIG3.4}).
367:
368: \begin{figure}[t]
369: \epsscale{1.}
370: \plotone{f5.eps}
371: \caption{Decomposition of the 30\,min average horizontal proper motions derived from LCT into speed
372: ($|v_{\rm LCT}|$, \textsl{left}) and azimuth maps ($\phi_{\rm LCT}$, \textsl{right}). The white or black contours outline the
373: umbral and penumbral boundaries of the sunspot. The white arrows in the
374: umbrae point towards the direction of disk center. The magnitude of the speed is
375: indicated by the color scale and the azimuth angle is indicated by the color-coded disk. }
376: \label{FIG3.6}
377: \end{figure}
378: \begin{figure*}[t]
379: \epsscale{1.1}
380: \plotone{f6.eps}
381: \caption{LOS dopplergram series corresponding to Figure~\ref{FIG3.4}
382: showing the Evershed flow in the
383: sunspot. Redshifts are positive corresponding to bright areas. The gray
384: scale in Panel c represents the magnitude of LOS velocity.}
385: \label{FIG3.7}
386: \end{figure*}
387:
388:
389: The average LCT proper motion vectors (see Figure~\ref{FIG3.3}) are decomposed
390: into two components, i.e., magnitude $|v_{\rm LCT}|$ and azimuthal
391: direction $\phi_{\rm LCT}$. They are displayed for each pixel in
392: Figure~\ref{FIG3.6}. From June\,7 to June\,11, when the sunspot has a
393: penumbra, the $\phi_{\rm LCT}$ component shows a radial inward motion
394: in the inner sunspot (including the umbra and the inner part of
395: penumbra) and a radial outward motion in the outer sunspot. The
396: outward motion extends beyond the visible boundary of the outer
397: penumbra for up to 5\arcsec. In other words, a 5\arcsec\-wide annular
398: zone showing a radial outward proper motion (collar flow) surrounds the penumbra (or the
399: sunspot). The $|v_{\rm LCT}|$ component shows that there are two
400: ring-like proper motion structures with large $|v_{\rm LCT}|$ values. One is
401: surrounding the penumbra-umbra interface corresponding to the inward
402: motion of penumbral grains, which has relatively small magnitude (with
403: average speeds of about 0.4\,km\,s$^{-1}$ and a maximum speed reaching
404: up to 0.8\,km\,s$^{-1}$). Its width is about 3\arcsec\
405: ($\sim$2200\,km). The other is surrounding the outer penumbra and
406: harbors the high speed portion of the outward motion of penumbral
407: structures and the immediate part of the moat zone. There the $|v_{\rm LCT}|$ component
408: exhibits a larger magnitude with an average speed of about
409: 0.8\,km\,s$^{-1}$ and a maximum speed reaching up to
410: 1.4\,km\,s$^{-1}$. Its width is about 5\arcsec\
411: ($\sim$3700\,km). \citet{november+etal1987} were the first to detect
412: a radial, large speed collar flow of granules from the penumbra into the immediate
413: surroundings from the flow field around sunspots
414: and named it ``annulus''. The ring-like motion patterns lose their symmetry when the
415: sunspot decays. Both inner and outer ring seem to be shifted towards
416: the limb-side for about 2\arcsec\ or 1500\,km. As can be seen in
417: Figure~\ref{FIG3.6}, the rings at the center-side occur at the
418: penumbra-granulation or penumbra-umbra boundary, while the rings at
419: the limb-side occur outside corresponding borders.
420:
421: Concentrating on the residual pore on June\,12, we see from the $\phi_{\rm LCT}$
422: component that there is still a confined region ($\sim$2\arcsec\ to 3\arcsec)
423: around the pore where we detect a horizontal proper motion towards the pore, which is also indicated by the
424: dividing line seen in Figure~\ref{FIG3.4}. Beyond the dividing line
425: $\phi_{\rm LCT}$ indicates a motion that is directed away from the pore.
426: The $|v_{\rm LCT}|$ component shows that both the inward and outward motion in and
427: around the pore are not as strong as those in and around the sunspot.
428:
429: The LOS dopplergrams are displayed in Figure~\ref{FIG3.7} for each day
430: corresponding to the arrangement in Figure~\ref{FIG3.4}. The dopplergrams
431: illustrate not only the dominant Evershed effect, but also the fine structure in
432: the LOS velocity of the whole sunspot. This includes an indication of
433: upflows (dark blueshifts) in umbral dots, the presence of penumbral flow channels,
434: and upflows in penumbral grains, which are more prominent
435: on the limb-side of the penumbra with a bright redshifted background. Furthermore,
436: the dopplergrams corroborate the results derived from LCT proper motion maps and the cork analysis.
437: In particular, the LOS flow field around the pore on June\,12 strongly suggests that the moat flow
438: is still present: on the limb side and center side of the pore we observe
439: a large patch of redshifts and blueshifts, respectively. At the large viewing angle
440: ($\mu = 0.43$), where also higher layers of the solar atmosphere
441: are probed, the moat flow component adds significantly
442: to the convective flow pattern in such a way that it increases the LOS
443: component of the granular upflows on the center side but overcompensates
444: the LOS component of the granular upflows limbwards of the pore. It is noticed that the dopplergrams for June 11 and 12 show much stronger LOS velocities far from the spot or pore than observations nearer disk center (e.g., June 7). The reason for that could be the convection of the supergranulation that contributes more to the LOS velocity towards the limb.
445:
446: From Figures~\ref{FIG3.2} and \ref{FIG3.4}, and by watching long-term intensity
447: and magnetogram movies from MDI, our visual impression is that magnetic flux was removed
448: from the sunspot by the following processes: fragmentation of the
449: spot in the north-eastern part and by MMFs.
450: To the west, the flux transported by those MMFs that have the same polarity as the spot
451: was promptly canceled by the leading negative flux
452: when the MMFs reached the outer edge of the moat region. To the east
453: where the network and plage showed much less opposite
454: polarity flux, substantial amount of flux
455: was transported by MMFs of the same polarity as the sunspot into the surrounding plage region of the same polarity.
456: Fragmentation, flux cancelation, and transport of flux to the
457: network and plage region were likely responsible for the decay of the sunspot.
458: On June\,12, the penumbra completely disappeared and only a pair
459: of pores remained. In addition, the sunspot had rotated counterclockwise by about
460: 45\arcdeg\ during the five days from June\,7 to June\,11.
461:
462: We also noticed that the relative brightness of both penumbra and umbra
463: increased when the sunspot moved from disk center to the limb. Moreover, this
464: brightness increase is faster in the umbra than in the penumbra. The decay of
465: the sunspot could be responsible for the overall increase of the sunspot
466: brightness. Another possibility is that the sunspot is relatively brighter from
467: the lower to the upper photosphere, which suggests that the temperature along
468: magnetic flux tube increases faster than that in weak or non-magnetic plasmas. This behavior of the thermal stratification agrees well with the cool umbral model of \citet{collados+etal1994}.
469:
470: To illustrate the decay process of the sunspot more quantitatively we
471: determined the change of the total sunspot area and the magnetic flux with time.
472: The result is shown in Fig.~\ref{FIG3.5}. The light and dark gray
473: vertical bars represent the penumbral and umbral areas, respectively. Their sum
474: corresponds to the total area of the sunspot. The areas were calculated using
475: the high-resolution data based on the enclosed areas shown by contours in
476: Figure~\ref{FIG3.4}. The total-to-umbral area ratio
477: ($r_{A} = A_{\rm tot}/A_{U}$) and the percentage of the umbral to total sunspot
478: area for each day were also calculated and annotated accordingly on the top and at
479: the bottom of each bar in Figure~\ref{FIG3.5}, respectively. The published
480: results of $r_{A}$ range from 4.0 to 6.3, but are mostly distributed between
481: 5.0 and 6.0 \citep[for a summary see][]{solanki2003}. The result from this study
482: shows that $r_{A}$ ranges from 5.19 to 6.31, which is in agreement with
483: previous measurements. The percentage of the umbral to total sunspot area
484: increased from 15.85\,\% to 18.98\,\% from June\,8 to June\,11, which indicates that the
485: sunspot penumbra decays faster than the umbra. The plus signs in
486: Figure~\ref{FIG3.5} show the evolution of the magnetic flux in the sunspot ($F
487: = \int B \cdot dS$) calculated from MDI magnetograms. The projection effect of
488: the area $S$ has been corrected and the magnetic fields $B$ were
489: partially corrected assuming that the field lines are horizontal in the penumbra and vertical in the umbra. From
490: June~9 to 12, the sunspot flux decayed with an almost constant decay rate of
491: $3.2 \times 10^{15}$\,Mx\,s$^{-1}$.
492:
493: \begin{figure}[t]
494: \epsscale{1.1}\vspace{0.5cm}
495: \plotone{f7.eps}
496: \caption{Evolution of sunspot area and magnetic flux.
497: The light gray bars represent the penumbral area,
498: dark gray bars correspond to the umbral area. Their sum (i.e., the whole bar)
499: relates to the total area of the sunspot. The series of plus signs shows
500: the evolution of the magnetic flux of the sunspot.
501: A linear fit is applied to the decaying
502: phase from June\,9 to 12. The slope provides the overall
503: decay rate of $3.2 \times 10^{15}$\,Mx\,s$^{-1}$.}
504: \label{FIG3.5}
505: \end{figure}
506:
507:
508:
509: \section{Discussion and Conclusion}\label{sec:discussion}
510:
511: A photometric and spectroscopic high-resolution study has been presented
512: showing the flow fields in a decaying, isolated, quasi-round sunspot covering
513: five days in June\,2005. Based on horizontal proper motions and LOS flow fields determined
514: from a LCT analysis and spectroscopic observations, respectively,
515: we confirm the existence of inward motions in the inner penumbra and
516: outward motions in the outer penumbra separated by a dividing line
517: located inside the penumbra \citep[see e.g.][]{zirin+wang1989, wang+zirin1992,
518: molownyhoras1994, denker1998, sobotka+brandt+simon1999, bovelet+wiehr2003}.
519: For the first time, we demonstrate that the dividing line survives the
520: decay process during which the sunspot lost its penumbra
521: and fragmented. Most interestingly, we find that the moat flow in the periphery of the
522: umbral core, i.e the residual pore, is still detectable after the penumbra has disappeared.
523: The outward motion, however, is not in the immediate surroundings of
524: the pore but separated by an annular inward motion, and is
525: much weaker than the one we find around the sunspot. We argue that the
526: residual pore retained the flow pattern bequeathed from its parent sunspot, since sunspots and pores have historic properties, i.e., they tend to continue their past physical properties, as evidenced by the fact that large pores can be larger than small sunspots \citep{solanki2003}.
527: The existence of a moat flow around umbral fragments or
528: individual pores is still debated in the literature
529: \citep[e.g. ][]{wang+zirin1992, yang+etal2003, vargasdominguez+etal2007}.
530: Generally, it is believed that the large-scale organized outflow pattern of
531: the moat around sunspots is driven by a temperature excess that builds
532: up beneath because normal convective energy transport is inhibited.
533: Furthermore, results from time-distance helioseismology corroborate
534: the existence of large-scale vortex cells and
535: downflows beneath sunspots \citep[e.g][]{duvall+etal1996, gizon+duvall+larsen2000,
536: zhao+kosovichev+duvall2001}. Already \citet{parker1992} suggested that individual
537: emerging flux bundles are collared by downdraft vortex rings.
538: Motivated by \cite{bovelet+wiehr2003} who put these ideas together illustrated
539: in a sketch (see their Figure~13) we explain our observation of
540: a collar flow directed towards the umbra after the penumbra has disappeared
541: as the signature of that downflow vortex cell, which is adjacent to an outer upflow and diverging vortex roll.
542:
543: The inward proper motions in the inner penumbra can be very well explained in the framework of the
544: moving-tube scenario by \cite{schlichenmaier+jahn+schmidt1998a}
545: mimicking the uncombed penumbra \citep{solanki+montavon1993}.
546: In this model the filamentary structure of the penumbra originates
547: from the dynamics of ascending flux tubes embedded in a static background
548: field. During the ascending phase a systematic flow develops along the flux tube,
549: the Evershed flow. In particular penumbral grains are interpreted as the upstream
550: footpoints of the flux tubes. As the flux tubes rise their cross-section
551: with the $\tau=1$-level migrates inwards giving the
552: visual impression that the penumbral grains move towards the umbra.
553: Many observations have provided support for this model like most recently
554: \citet{rimmele+marino2006}. The outward migration detected in the outer
555: penumbra might be explained by the moving-tube model if the photospheric sea-serpent structure developed from the penumbral flux tubes simulated by \citet{schlichenmaier2002} occurs in the real Sun. We speculate that the outward flow which extends beyond the visible boundary of the sunspot penumbra is related to the propagation of ECs \citep[see][]{cabrera+etal2007}
556: but this needs still to be verified.
557:
558: The observed sunspot was a peculiar follower in the sense that it was
559: larger, more regular and more stable than its leading spot. \citet{martinezpillet2002}
560: reviewed several aspects related to the decay of leading and following sunspots
561: and summarized the current state of knowledge of the physical mechanisms that
562: are involved in sunspot evolution. In general, a long period of stability is
563: rarely found in followers, they decay very fast, typically
564: within several days or less after their formation. Following spots usually
565: possess irregular shapes and have incomplete penumbrae. Only 3\,\% of the observed
566: followers develop a round shape and appear in the so-called $\alpha$ configuration
567: \citep{bray+loughhead1964}. It is much more likely for the preceding sunspot to
568: develop such a configuration which then can last from days to
569: weeks. Hence, the observations presented in this study provide a rare
570: exception of the aforementioned rules. We find that during the decay
571: phase the percentage of the umbral to total sunspot area
572: increased from 15.9\,\% to 19.0\,\% indicating that the penumbra decays
573: faster than the umbra. Although the flux decay rate was almost constant
574: ($3.2 \times 10^{15}$\,Mx\,s$^{-1}$), the decay process was not uniform.
575: We observed that the decay proceeded by the following mechanisms:
576: fragmentation of the sunspot, flux cancelation of MMFs (of the same polarity as the sunspot)
577: that encounter the leading opposite polarity network and plages areas, and flux transport by MMFs (of the same polarity as the sunspot)
578: to the network and plage regions that have the same polarity of the sunspot.
579:
580: In order to reconcile the discrepancies found in the large-scale flow patterns
581: surrounding sunspots, umbral fragments, individual pores that never developed
582: a penumbra (innate pores), and individual pores that remained from a decayed sunspot (residual pores) not only better statistics is needed. To fully understand
583: the connection between the flow patterns and the evolution of the magnetic structures
584: high-spatial resolution spectropolarimetric observations as provided
585: by HINODE must complete spectroscopic and imaging observations like those
586: presented in this investigation.
587:
588:
589: \acknowledgments
590: We would like to thank John Thomas, Valantina Abramenco, and Dale E. Gary
591: for comments and discussions on the original manuscript. We also thank the referee for a number of valuable suggestions that help to improve the paper.
592: This research has made use of NASA's Astrophysics Data System (ADS).
593: This work was supported by NSF under grant ATM 03-42560, ATM 03-13591,
594: ATM 02-36945, ATM 05-48952, and MRI AST 00-79482 and by NASA under grant NAG 5-13661.
595:
596:
597: \begin{thebibliography}{49}
598:
599: \bibitem[{{Bovelet} \& {Wiehr}(2003)}]{bovelet+wiehr2003}
600: {Bovelet}, B., \& {Wiehr}, E. 2003, \aap, 412, 249
601:
602: \bibitem[{{Bray} \& {Loughhead}(1964)}]{bray+loughhead1964}
603: {Bray}, R.~J., \& {Loughhead}, R.~E. 1964, {Sunspots} (The International
604: Astrophysics Series, London: Chapman \& Hall, 1964)
605:
606: \bibitem[{{Brickhouse} \& {Labonte}(1988)}]{brickhouse+labonte1988}
607: {Brickhouse}, N.~S., \& {Labonte}, B.~J. 1988, \solphys, 115, 43
608:
609: \bibitem[{{Cabrera Solana} {et~al.}(2006){Cabrera Solana}, {Bellot Rubio},
610: {Beck}, \& {del Toro Iniesta}}]{cabrera+etal2006a}
611: {Cabrera Solana}, D., {Bellot Rubio}, L.~R., {Beck}, C., \& {del Toro Iniesta},
612: J.~C. 2006, \apjl, 649, L41
613:
614: \bibitem[{{Cabrera Solana} {et~al.}(2007){Cabrera Solana}, {Bellot Rubio},
615: {Beck}, \& {del Toro Iniesta}}]{cabrera+etal2007}
616: ---. 2007, ArXiv e-prints, 707
617:
618: \bibitem[{{Chapman} {et~al.}(2003){Chapman}, {Dobias}, {Preminger}, \&
619: {Walton}}]{chapman+etal2003}
620: {Chapman}, G.~A., {Dobias}, J.~J., {Preminger}, D.~G., \& {Walton}, S.~R. 2003,
621: \grl, 30, 27
622:
623: \bibitem[{{Choudhary} \& {Balasubramaniam}(2007)}]{choudhary+bala2007}
624: {Choudhary}, D.~P., \& {Balasubramaniam}, K.~S. 2007, \apj, 664, 1228
625:
626: \bibitem[{{Collados} {et~al.}(1994){Collados}, {Martinez}, {Ruiz}, {Deltoro}, \& {Vasquez}}]{collados+etal1994}
627: {Collados}, M., {Martinez}, V., {Ruiz}, B., {Deltoro}, J.~C., \& {Vasquez}, M. 1994,
628: AA, 0,1
629:
630: \bibitem[{{Denker}(1998)}]{denker1998}
631: {Denker}, C. 1998, \solphys, 180, 81
632:
633: \bibitem[{{Denker} \& {Tritschler}(2005)}]{denker+tritschler2005}
634: {Denker}, C., \& {Tritschler}, A. 2005, \pasp, 117, 1435
635:
636: \bibitem[{{Denker} {et~al.}(2005){Denker}, {Mascarinas}, {Xu}, {Cao}, {Yang},
637: {Wang}, {Goode}, \& {Rimmele}}]{denker+etal2005}
638: {Denker}, C., {Mascarinas}, D., {Xu}, Y., {Cao}, W., {Yang}, G., {Wang}, H.,
639: {Goode}, P.~R., \& {Rimmele}, T. 2005, \solphys, 227, 217
640:
641: \bibitem[{{Denker} {et~al.}(2007{\natexlab{a}}){Denker}, {Deng}, {Rimmele},
642: {Tritschler}, \& {Verdoni}}]{denker+etal2007b}
643: {Denker}, C., {Deng}, N., {Rimmele}, T.~R., {Tritschler}, A., \& {Verdoni}, A.
644: 2007{\natexlab{a}}, \solphys, 241, 411
645:
646: \bibitem[{{Denker} {et~al.}(2007{\natexlab{b}}){Denker}, {Tritschler},
647: {Rimmele}, {Richards}, {Hegwer}, \& {W{\"o}ger}}]{denker+etal2007a}
648: {Denker}, C., {Tritschler}, A., {Rimmele}, T.~R., {Richards}, K., {Hegwer},
649: S.~L., \& {W{\"o}ger}, F. 2007{\natexlab{b}}, \pasp, 119, 170
650:
651: \bibitem[{{Duvall} {et~al.}(1996){Duvall}, {D'Silva}, {Jefferies}, {Harvey}, \&
652: {Schou}}]{duvall+etal1996}
653: {Duvall}, T.~L.~J., {D'Silva}, S., {Jefferies}, S.~M., {Harvey}, J.~W., \&
654: {Schou}, J. 1996, \nat, 379, 235
655:
656: \bibitem[{{Evershed}(1909)}]{evershed1909}
657: {Evershed}, J. 1909, \mnras, 69, 454
658:
659: \bibitem[{{Gizon} {et~al.}(2000){Gizon}, {Duvall}, \&
660: {Larsen}}]{gizon+duvall+larsen2000}
661: {Gizon}, L., {Duvall}, Jr., T.~L., \& {Larsen}, R.~M. 2000, Journal of
662: Astrophysics and Astronomy, 21, 339
663:
664: \bibitem[{{Hagenaar} \& {Shine}(2005)}]{hagenaar+shine2005b}
665: {Hagenaar}, H.~J., \& {Shine}, R.~A. 2005, \apj, 635, 659
666:
667: \bibitem[{{Harvey} \& {Harvey}(1973)}]{harvey+harvey1973}
668: {Harvey}, K., \& {Harvey}, J. 1973, \solphys, 28, 61
669:
670: \bibitem[{{Kubo} {et~al.}(2007){Kubo}, {Shimizu}, \&
671: {Tsuneta}}]{kubo+shimizu+tsuneta2007}
672: {Kubo}, M., {Shimizu}, T., \& {Tsuneta}, S. 2007, \apj, 659, 812
673:
674: \bibitem[{{Leka} \& {Skumanich}(1998)}]{leka+skumanich1998}
675: {Leka}, K.~D., \& {Skumanich}, A. 1998, \apj, 507, 454
676:
677: \bibitem[{{Mart{\'{\i}}nez Pillet}(2002)}]{martinezpillet2002}
678: {Mart{\'{\i}}nez Pillet}, V. 2002, Astronomische Nachrichten, 323, 342
679:
680: \bibitem[{{Mart{\'{\i}}nez Pillet} {et~al.}(1993){Martinez Pillet}, {Moreno-Insertis},
681: \& {Vazquez}}]{martinezpillet+morenoinsertis+vazquez1993}
682: {Martinez Pillet}, V., {Moreno-Insertis}, F., \& {Vazquez}, M. 1993, \aap, 274,
683: 521
684:
685: \bibitem[{{McIntosh}(1981)}]{mcintosh1981}
686: {McIntosh}, P.~S. 1981, in The Physics of Sunspots, ed. L.~E. {Cram} \& J.~H.
687: {Thomas}, 7--54
688:
689: \bibitem[{{Molowny-Horas}(1994)}]{molownyhoras1994}
690: {Molowny-Horas}, R. 1994, \solphys, 154, 29
691:
692: \bibitem[{{Moreno-Insertis} \& {Vazquez}(1988)}]{morenoinsertis+vazquez1988}
693: {Moreno-Insertis}, F., \& {Vazquez}, M. 1988, \aap, 205, 289
694:
695: \bibitem[{{November}(1986)}]{november1986}
696: {November}, L.~J. 1986, \ao, 25, 392
697:
698: \bibitem[{{November} {et~al.}(1986){November}, {Simon}, {Tarbell}, \&
699: {Title}}]{november+etal1986}
700: {November}, L.~J., {Simon}, G.~W., {Tarbell}, T.~D., \& {Title}, A.~M. 1986, in
701: Bulletin of the American Astronomical Society, Vol.~18, Bulletin of the
702: American Astronomical Society, 665
703:
704: \bibitem[{{November} {et~al.}(1987){November}, {Simon}, {Tarbell}, {Title}, \&
705: {Ferguson}}]{november+etal1987}
706: {November}, L.~J., {Simon}, G.~W., {Tarbell}, T.~D., {Title}, A.~M., \&
707: {Ferguson}, S.~H. 1987, in Theoretical Problems in High Resolution Solar
708: Physics, ed. G.~{Athay} \& D.~S. {Spicer}, 121--127
709:
710: \bibitem[{{Parker}(1992)}]{parker1992}
711: {Parker}, E.~N. 1992, \apj, 390, 290
712:
713: \bibitem[{{Petrovay} {et~al.}(1999){Petrovay}, {Mart{\'{\i}}nez Pillet}, \&
714: {van Driel-Gesztelyi}}]{petrovay+martinezpillet+vandrielgesztelyi1999}
715: {Petrovay}, K., {Mart{\'{\i}}nez Pillet}, V., \& {van Driel-Gesztelyi}, L.
716: 1999, \solphys, 188, 315
717:
718: \bibitem[{{Ravindra}(2006)}]{ravindra2006}
719: {Ravindra}, B. 2006, \solphys, 237, 297
720:
721: \bibitem[{{Rimmele} \& {Marino}(2006)}]{rimmele+marino2006}
722: {Rimmele}, T., \& {Marino}, J. 2006, \apj, 646, 593
723:
724: \bibitem[{{Rimmele}(1994)}]{rimmele1994}
725: {Rimmele}, T.~R. 1994, \aap, 290, 972
726:
727: \bibitem[{{Rimmele} {et~al.}(2004){Rimmele}, {Richards}, {Hegwer}, {Fletcher},
728: {Gregory}, {Moretto}, {Didkovsky}, {Denker}, {Dolgushin}, {Goode},
729: {Langlois}, {Marino}, \& {Marquette}}]{rimmele+etal2004}
730: {Rimmele}, T.~R., {Richards}, K., {Hegwer}, S., {Fletcher}, S., {Gregory}, S.,
731: {Moretto}, G., et~al. 2004, Proceedings SPIE, 5171, 179-186
732:
733: \bibitem[{{Rouppe van der Voort}(2003)}]{rouppe2003}
734: {Rouppe van der Voort}, L.~H.~M. 2003, \aap, 397, 757
735:
736: \bibitem[{{Sainz Dalda} \& {Mart{\'{\i}}nez
737: Pillet}(2005)}]{sainzdalda+martinezpillet2005}
738: {Sainz Dalda}, A., \& {Mart{\'{\i}}nez Pillet}, V. 2005, \apj, 632, 1176
739:
740: \bibitem[{{Spirock}(2005)}]{spirock2005}
741: {Spirock}, T. 2005, PhD Thesis of New Jersey Insititute of Technology
742:
743: \bibitem[{{Scherrer} {et~al.}(1995){Scherrer}, {Bogart}, {Bush}, {Hoeksema},
744: {Kosovichev}, {Schou}, {Rosenberg}, {Springer}, {Tarbell}, {Title},
745: {Wolfson}, {Zayer}, \& {MDI Engineering Team}}]{scherrer+etal1995}
746: {Scherrer}, P.~H., {Bogart}, R.~S., {Bush}, R.~I., {Hoeksema}, J.~T.,
747: {Kosovichev}, A.~G., {Schou}, J., et~al. 1995, \solphys, 162, 129
748:
749: \bibitem[{{Schlichenmaier} {et~al.}(1998){Schlichenmaier}, {Jahn}, \&
750: {Schmidt}}]{schlichenmaier+jahn+schmidt1998a}
751: {Schlichenmaier}, R., {Jahn}, K., \& {Schmidt}, H.~U. 1998, \apjl, 493, L121
752:
753: \bibitem[{{Schlichenmaier}(2002){Schlichenmaier}}]{schlichenmaier2002}
754: {Schlichenmaier}, R. 2002, Astronomische Nachrichten, 323, 303
755:
756: \bibitem[{{Schmidt} {et~al.}(1999){Schmidt}, {Stix}, \&
757: {W{\"o}hl}}]{schmidt+stix+woehl1999}
758: {Schmidt}, W., {Stix}, M., \& {W{\"o}hl}, H. 1999, \aap, 346, 633
759:
760: \bibitem[{{Sheeley}(1969)}]{sheeley1969}
761: {Sheeley}, Jr., N.~R. 1969, \solphys, 9, 347
762:
763: \bibitem[{{Shine} {et~al.}(1994){Shine}, {Title}, {Tarbell}, {Smith}, {Frank},
764: \& {Scharmer}}]{shine+etal1994}
765: {Shine}, R.~A., {Title}, A.~M., {Tarbell}, T.~D., {Smith}, K., {Frank}, Z.~A.,
766: \& {Scharmer}, G. 1994, \apj, 430, 413
767:
768: \bibitem[{{Sobotka} {et~al.}(1999){Sobotka}, {Brandt}, \&
769: {Simon}}]{sobotka+brandt+simon1999}
770: {Sobotka}, M., {Brandt}, P.~N., \& {Simon}, G.~W. 1999, \aap, 348, 621
771:
772: \bibitem[{{Solanki}(2003)}]{solanki2003}
773: {Solanki}, S.~K. 2003, Astron.\ Astrophys.\ Rev.\, 11, 153
774:
775: \bibitem[{{Solanki} \& {Montavon}(1993)}]{solanki+montavon1993}
776: {Solanki}, S.~K., \& {Montavon}, C. A.~P. 1993, \aap, 275, 283
777:
778: \bibitem[{{Vargas Dom{\'{\i}}nguez} {et~al.}(2007){Vargas Dom{\'{\i}}nguez},
779: {Bonet}, {Mart{\'{\i}}nez Pillet}, {Katsukawa}, {Kitakoshi}, \& {Rouppe van
780: der Voort}}]{vargasdominguez+etal2007}
781: {Vargas Dom{\'{\i}}nguez}, S., {Bonet}, J.~A., {Mart{\'{\i}}nez Pillet}, V.,
782: {Katsukawa}, Y., {Kitakoshi}, Y., \& {Rouppe van der Voort}, L. 2007, \apjl,
783: 660, L165
784:
785: \bibitem[{{Wang} \& {Zirin}(1992)}]{wang+zirin1992}
786: {Wang}, H., \& {Zirin}, H. 1992, \solphys, 140, 41
787:
788: \bibitem[{{Wang} {et~al.}(1998){Wang}, {Denker}, {Spirock}, {Goode}
789: {Yang}, {Marquette}, {Varsik}, {Fear},
790: {Nenow}, \& {Dingley}}]{wang+etal1998}
791: {Wang}, H., {Denker}, C., {Spirock}, T., {Goode}, P.~R.,
792: {Yang}, S., {Marquette}, W., {Varsik}, J., {Fear}, R.~J.,
793: {Nenow}, J. \& {Dingley}, D.~D. 1998, \solphys, 183, 1
794:
795: \bibitem[{{Yang} {et~al.}(2003){Yang}, {Xu}, {Wang}, \&
796: {Denker}}]{yang+etal2003}
797: {Yang}, G., {Xu}, Y., {Wang}, H., \& {Denker}, C. 2003, \apj, 597, 1190
798:
799: \bibitem[{{Yurchyshyn} \& {Wang}(2001)}]{yurchyshyn+wang2001}
800: {Yurchyshyn}, V.~B., \& {Wang}, H. 2001, \solphys, 203, 233
801:
802: \bibitem[{{Zhao} {et~al.}(2001){Zhao}, {Kosovichev}, \&
803: {Duvall}}]{zhao+kosovichev+duvall2001}
804: {Zhao}, J., {Kosovichev}, A.~G., \& {Duvall}, Jr., T.~L. 2001, \apj, 557, 384
805:
806: \bibitem[{{Zirin} \& {Wang}(1989)}]{zirin+wang1989}
807: {Zirin}, H., \& {Wang}, H. 1989, \solphys, 119, 245
808:
809: \bibitem[{{Zwaan}(1992)}]{zwaan1992}
810: {Zwaan}, C. 1992, in NATO ASIC Proc. 375: Sunspots. Theory and Observations,
811: ed. J.~H. {Thomas} \& N.~O. {Weiss}, 75--100
812:
813: \end{thebibliography}
814:
815:
816:
817:
818: \end{document}
819:
820:
821:
822:
823:
824: