1: % mn2esample.tex
2: %
3: % v2.1 released 22nd May 2002 (G. Hutton)
4:
5: \documentclass[useAMS,usenatbib]{mn2e}
6: \usepackage{graphicx}
7: % If your system does not have the AMS fonts version 2.0 installed, then
8: % remove the useAMS option.
9: %
10: % useAMS allows you to obtain upright Greek characters.
11: % e.g. \umu, \upi etc. See the section on "Upright Greek characters" in
12: % this guide for further information.
13: %
14: % If you are using AMS 2.0 fonts, bold math letters/symbols are available
15: % at a larger range of sizes for NFSS release 1 and 2 (using \boldmath or
16: % preferably \bmath).
17: %
18: % The usenatbib command allows the use of Patrick Daly's natbib.sty for
19: % cross-referencing.
20: %
21: % If you wish to typeset the paper in Times font (if you do not have the
22: % PostScript Type 1 Computer Modern fonts you will need to do this to get
23: % smoother fonts in a PDF file) then uncomment the next line
24: % \usepackage{Times}
25:
26: %%%%% AUTHORS - PLACE YOUR OWN MACROS HERE %%%%%
27:
28:
29: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
30:
31: \title[The solar flare-induced sunquake of 2005 January 15]{Helioseismic analysis of the solar flare-induced sunquake of 2005 January 15}
32: \author[H. Moradi, A. -C. Donea, C. Lindsey, D. Besliu-Ionescu and P. S. Cally]{H. Moradi$^{1}$\thanks{E-mail:
33: hamed.moradi@sci.monash.edu.au}, A. -C. Donea$^1$, C. Lindsey$^2$, D. Besliu-Ionescu$^{1,3}$ and P. S. Cally$^1$\\
34: $^{1}$Centre for Stellar and Planetary Astrophysics, School of
35: Mathematical Sciences, Monash University, Victoria 3800, Australia\\
36: $^{2}$Colorado Research Associates Division, NorthWest Research
37: Associates Inc., 3380 Mitchell Lane, Boulder, CO, 80301, U.S.A\\
38: $^{3}$Astronomical Institute of the Romanian Academy, RO-040557,
39: Bucharest, Romania\\}
40:
41: \begin{document}
42:
43: %\date{Submitted}
44:
45:
46: \maketitle
47:
48:
49: \begin{abstract}
50: We report the discovery of one of the most powerful sunquakes
51: detected to date, produced by an X1.2-class solar flare in active
52: region 10720 on 2005 January 15. We used helioseismic holography to
53: image the source of seismic waves emitted into the solar interior from
54: the site of the flare. Acoustic egression power maps at 3 and 6 mHz
55: with a 2~mHz bandpass reveal a compact acoustic source strongly
56: correlated with impulsive hard X-ray and visible-continuum emission
57: along the penumbral neutral line separating the two major opposing
58: umbrae in the $\delta$-configuration sunspot that predominates
59: AR10720. At 6~mHz the seismic source has two components, an intense,
60: compact kernel located on the penumbral neutral line of the
61: $\delta$-configuration sunspot that predominates AR10720, and a
62: significantly more diffuse signature distributed along the neutral
63: line up to $\sim$15~Mm east and $\sim$30~Mm west of the kernel. The
64: acoustic emission signatures were directly aligned with both hard
65: X-ray and visible continuum emission that emanated during the flare.
66: The visible continuum emission is estimated at $2.0 \times 10^{23}$~J,
67: approximately 500 times the seismic emission of $\sim 4 \times
68: 10^{20}$~J. The flare of 2005 January 15 exhibits the same close
69: spatial alignment between the sources of the seismic emission and
70: impulsive visible continuum emission as previous flares, reinforcing
71: the hypothesis that the acoustic emission may be driven by heating of
72: the low photosphere. However, it is a major exception in that there
73: was no signature to indicate the inclusion of protons in the particle
74: beams thought to supply the energy radiated by the flare. The
75: continued strong coincidence between the sources of seismic emission
76: and impulsive visible continuum emission in the case of a
77: proton-deficient white-light flare lends substantial support to the
78: ``back\,--\,warming'' hypothesis, that the low photosphere is
79: significantly heated by intense Balmer and Paschen continuum-edge
80: radiation from the overlying chromosphere in white-light flares.
81:
82:
83: \end{abstract}
84:
85: \begin{keywords}
86:
87: Sun: helioseismology -- Sun: flares -- Sun: oscillations
88:
89: \end{keywords}
90:
91: \section{Introduction}
92:
93: Although most large solar flares appear to be acoustically inactive,
94: certain energetic flares radiate intense seismic transients into the
95: solar interior during the impulsive phase. These wave packets radiate
96: thousands of kilometres from the flaring region into the solar
97: interior, but most of this energy is refracted back to the solar
98: surface within approximately 50~Mm of the source and within an hour of
99: the beginning of the flare. The surface manifestation is a
100: wave-packet of ripples accelerating outward from the general source
101: region that is sometimes obvious in raw helioseismic observations.
102: \citet{kz1998} discovered the first known instance of seismic
103: emission, from the X2-class flare of 1996 July 9 in AR7978,
104: identifying the phenomenon by the name ``sunquake.''
105:
106: For a long time these events were thought to be an extremely rare phenomenon.
107: However, with the advancement of local helioseismic techniques such as
108: helioseismic holography \citep{lb2000}, we have now detected numerous seismic
109: sources of varying intensity produced by X- and high M-class flares
110: \citep{dbl1999,dl2005,donea2005}.
111:
112: A subsequent extensive survey of X-class solar flares (Besliu-Ionescu et al., in preparation)
113: led to the discovery of more than a dozen seismic emission signatures
114: from flares. Almost all of these have occurred in complex active
115: regions. In this paper we report on the discovery of one of the most
116: powerful flare seismic transients detected to date and compare the
117: acoustic signatures of this sunquake with other supporting
118: observations.
119:
120: AR10720 was a complex active region that appeared on the solar disk on
121: 2005 January 11 and soon became one of the largest and most active
122: sunspot regions of the current solar cycle. In the period January 15 -- 20,
123: AR10720 produced 5 X-class solar flares, including an X7.1 on
124: January 20, which produced an intense solar proton storm. However,
125: helioseismic observations sufficient to show seismic emission were
126: acquired only for the X1.2 flare of January 15. This flare was
127: observed by numerous space and ground-based solar observatories,
128: including the Michelson Doppler Imager (MDI) instrument on board the Solar and Heliospheric Observatory (\textit{SOHO}), the Reuven
129: Ramaty High-Energy Solar Spectroscopic Imager (\textit{RHESSI}), the
130: Geostationary Operational Environmental Satellite (\textit{GOES}),
131: the Transition Region and Coronal Explorer (\textit{TRACE}), and the
132: earth-based Global Oscillations Network Group (GONG).
133: AR10720 itself was observed by the Imaging Vector Magnetograph (IVM)
134: at the Mees Solar Observatory in the general time frame of the 15
135: January 2005 flare.
136:
137: In Section 2 we present a brief description of the helioseismic observations
138: we analyzed and in Section 3 we review the technique of computational seismic
139: holography.
140: Section 4 consists of our results and analysis, and finally in Section 5 we
141: present a discussion and summary.
142:
143: \section[]{The Helioseismic Observations}
144:
145: The MDI data consist of full-disk Doppler images in the photospheric line
146: Ni~I~6768~\AA, obtained at a cadence of 1 minute, in addition to approximately
147: hourly continuum intensity images and line-of-sight magnetograms.
148: The MDI data sets are described in more detail by \citet{setal1995}.
149: For the flare of 2005 January 15, we analyzed a dataset with a period of
150: four hours encompassing the flare.
151: For the purpose of our analysis, the MDI images we obtained
152: (Dopplergrams, magnetograms and intensity continuum) were remapped onto
153: a Postel projection \citep{deforest2004} that tracks solar rotation,
154: with the region of interest fixed at the centre of the projection.
155: The nominal pixel separation of the projection was 0.002 solar radii
156: (1.4~Mm) with a $256\times256$ pixels field of view, thus encompassing
157: a region of approximately $360 \times 360$~Mm$^2$ on the solar surface.
158:
159: \section{Helioseismic Holography}
160:
161: We briefly review the adaptation of computational seismic holography
162: for applications in flare seismology. In general, helioseismic
163: holography is the phase-coherent reconstruction of acoustic waves
164: observed at the solar surface into the solar interior to render
165: stigmatic images of subsurface sources that have given rise to the
166: surface disturbance. Because the solar interior refracts down-going
167: waves back to the surface, helioseismic holography can likewise use
168: observations in one surface region, the pupil, to image another, the
169: focus, a considerable distance from the pupil. We call this ``seismic
170: holography from the subjacent vantage'' (see fig. 4 of
171: \citet{lb2000}). The subjacent vantage renders the photosphere
172: as viewed by an acoustic observer directly beneath it.
173:
174: In general the acoustic reconstruction can be done either forward or
175: backward in time. When it is backward in time, we call the
176: extrapolated field the ``acoustic egression.'' In the case of
177: subjacent vantage holography, this represents waves emanating from the
178: surface focus downward into the solar interior. When the surface
179: acoustic field at any point $\textbf{r}'$ in the pupil is expressed as
180: a complex amplitude $\hat{\psi}$ for any given frequency $\omega$, the
181: acoustic egression can be expressed as
182: \begin{equation}
183: \hat H_+(\textbf{r}, ~\omega)
184: ~=~ \int_{pupil} \hat{G}_+(\textbf{r}, ~\textbf{r}', ~\omega)
185: ~\hat{\psi}(\textbf{r}', ~\omega)d^2\textbf{r}'.
186: \label{eq:Hpluscalc}
187: \end{equation}
188:
189: In this formalism,
190: $\hat{G}_+(\textbf{r}, ~\textbf{r}',~\omega)$ is a Green's function that
191: expresses the disturbance at the focus, $\textbf{r}$, due to a measured point
192: source at surface point $\textbf{r}'$ from which the acoustic wave is supposed
193: to propagate backwards in time to the focus.
194:
195: The relation between the complex amplitude, $\hat\psi(\textbf{r},~\omega)$,
196: of frequency appearing in equation (\ref{eq:Hpluscalc}) and the real acoustic
197: field, $\psi(\textbf{r},~t)$, representing the surface acoustic field in the
198: MDI observations as a function of time is expressed by the Fourier transform:
199: \begin{equation}
200: \psi(\textbf{r}, ~t)
201: ~=~ \frac{1}{\sqrt{2\pi}}\int^{\infty}_{-\infty} e^{i\omega t}
202: ~\hat{\psi}(\textbf{r}, ~\omega) ~d\omega.
203: \label{eq:psit}
204: \end{equation}
205: The same applies to the acoustic egression:
206: \begin{equation}
207: H_+(\textbf{r}, ~t)
208: ~=~ \frac{1}{\sqrt{2\pi}}\int^{\infty}_{-\infty}e^{iwt}
209: ~\hat{H}_+(\textbf{r}, ~\omega) ~d\omega.
210: \label{eq:Hplust}
211: \end{equation}
212: The ``egression power",
213: \begin{equation}
214: P(\textbf{r}, ~t) ~=~ |H_+(\textbf{r}, ~t)|^2,
215: \label{eq:Hpwr}
216: \end{equation}
217: is used extensively in holography of acoustic sources and absorbers.
218: Equation (\ref{eq:Hpwr}) is used to produce egression power maps, which show
219: compact positive signatures in the spatial and temporal neighbourhoods of
220: localized seismic transient emitters.
221: The signature of a localized absorber illuminated by ambient acoustic noise
222: is a similarly sharp deficit in egression power, appearing as a silhouette
223: against a generally positive background when rendered graphically.
224:
225: In this study $P({\bf r}, t)$ is separately derived from computations
226: of $\hat H_+({\bf r}, \omega )$ over 2\,--\,4~mHz and 5\,--\,7~mHz
227: ranges of the spectrum. In practice, there are major diagnostic
228: advantages to the 5\,--\,7~mHz spectrum, as it avoids the much greater
229: quiet-Sun ambient noise at lower frequencies, which competes
230: unfavourably with acoustic emission into the pupil from the flare. Due
231: to a shorter wavelength, the high frequency band also provides us with
232: waves that have a finer diffraction limit. These advantages come at
233: some expense in temporal discrimination, as the egression power
234: signatures that result are temporally smeared to a minimum effective
235: duration of order
236: \begin{equation}
237: \Delta t ~=~ \frac{1}{\Delta\nu} ~=~ \frac{1}{2~{\rm mHz}} ~=~ 500~{\rm s}.
238: \label{eq:Deltat}
239: \end{equation}
240: This smearing operates in both directions in time, meaning that the acoustic
241: signature of the flare $P({\bf r}, t)$, once the computation is complete, will
242: invariably commence several minutes before the actual onset of the flare and
243: last for several minutes afterward even if the actual acoustic disturbance was
244: instantaneous (and no seismic signature at all reached the pupil until nearly
245: twenty minutes after the flare began).
246: \begin{figure*}
247: \begin{center}
248: %\includegraphics[width=0.8\textwidth]{images/egress_pwr_snaps.eps}
249: \includegraphics[width=0.7\textwidth]{figure1.eps}
250: \caption{Egression power snapshots of AR10720 on 2005 January 15 taken
251: before, during and after the flare and integrated over a 5\,--\,7~mHz
252: frequency band. The top frames show an MDI visible continuum image of
253: AR10720 (left) at 00:00~UT and a magnetogram (right) at 00:28~UT.
254: The second row shows continuum intensity differences 30 seconds before and
255: after the time that appears above the respective frames, taken from
256: the GONG observatory at Mauna Loa. The bottom three rows show egression
257: power maps before (row 3), during (row 4), and after (bottom row) the
258: flare at 3~mHz (left column) and 6~mHz (right column). The annular
259: pupil for the egression computations is drawn in the top left panel.
260: To improve statistics, the original egression power snapshots are
261: smeared by convolution with a Gaussian with a $1/e$-half-width of
262: 3~Mm. Times are indicated above respective panels, with arrows
263: inserted to indicate the location of the acoustic source. Colour
264: scales at right and left of row 3 apply to respective columns in rows
265: 3\,--\,5. Egression power images and the continuum images are
266: normalized to unity at respective mean quiet-Sun values. At 3~mHz
267: this is $\sim$2~kW~m$^{-2}$. At 6~mHz it is 70~W~m$^{-2}$.}
268: \label{fig:egpwrsnaps}
269: \end{center}
270: \end{figure*}
271:
272:
273: \section{Results and Analysis}
274:
275: \subsection{The Seismic Signatures}
276:
277: AR10720 was predominated by a single
278: $\delta$-configuration sunspot. The top row of
279: Fig.~\ref{fig:egpwrsnaps} shows continuum intensity (left) and
280: line-of-sight magnetic field (right) of the active region shortly
281: before the flare. The 2005 January 15 solar flare in AR10720 was
282: classified as X1.2, localized at N14E08 on the solar surface. The
283: \textit{GOES} satellite measured a $1.2\times10^{-1}$Jm$^{-2}$ X-ray flux in
284: the 1-8~\AA~range integrated over the duration of the flare. Excess
285: X-ray emission began at 00:22~UT, reaching a maximum at 00:43~UT, and
286: ending at 01:02~UT. There was significant white-light emission
287: with a sudden onset, as indicated by the intensity difference
288: signatures shown in the second row of Fig.~\ref{fig:egpwrsnaps}, and
289: this coincided closely with hard X-ray (HXR) signatures indicating high-energy
290: particles accelerated into the chromosphere. However, unlike the
291: flares of 2003 October 28\,--\,29 \citep{dl2005}, there were no
292: signatures to indicate the inclusion of high-energy protons in these
293: particle influxes.
294: \begin{figure*}
295: \begin{center}
296: %\includegraphics[width=0.8\textwidth, angle=0]{images/local_ac_pwr_snaps.eps}
297: \includegraphics[width=0.7\textwidth, angle=0]{figure2.eps}
298: \caption{Acoustic power snapshots of AR10720 on 2005 January 15.
299: Details are the same as for Fig.~\ref{fig:egpwrsnaps}, but local acoustic power maps
300: appear in the bottom three rows in place of egression power maps.}
301: \label{fig:acpwrsnaps}
302: \end{center}
303: \end{figure*}
304:
305: To assess seismic emission from the flare, we computed the egression,
306: $H_+$, as prescribed by equation~(\ref{eq:Hplust}) over the
307: neighbourhood of the active region at one-minute intervals in $t$,
308: mapping the egression power, $P$, as prescribed by
309: equation~(\ref{eq:Hpwr}), for each value of $t$. We call a map of $P$
310: evaluated at any single $t$ an egression power ``snapshot.'' From
311: this point will refer to the 5\,--\,7~mHz bandpass simply as 6~mHz and
312: to the 2\,--\,4~mHz bandpass as 3~mHz. Egression power snapshots
313: before, during and after the flare are shown in the bottom three rows
314: of Fig.~\ref{fig:egpwrsnaps} at 3~mHz (left column) and 6~mHz (right
315: column). In these computations the pupil was an annulus of radial
316: range 15\,--\,45~Mm centred on the focus
317: (Fig.~\ref{fig:egpwrsnaps}a).
318:
319: All egression power snapshots mapped in Fig.~{\ref{fig:egpwrsnaps}} show
320: considerably suppressed acoustic emission from the magnetic region, attributed to
321: strong acoustic absorption by magnetic regions, discovered by \citet{bdl1988} (see also \citet{braun1995,blff1998,bl1999a}). Furthermore, all 6~mHz egression
322: power snapshots in Fig.~{\ref{fig:acpwrsnaps}} also show acoustic emission ``halos,'' i.e. significantly
323: enhanced acoustic emission from the outskirts of complex active regions
324: \citep{bl1999b,dbl1999}.
325: %\footnote{\citet{bl1999b} and \citet{dbl1999} found
326: %conspicuous high-frequency acoustic emission halos surrounding all large,
327: %magnetically complex active regions.
328: %In fact, the outskirts of isolated, monopolar sunspots showed measurably enhanced
329: %acoustic emission \citep{lb1999} but this was rather subtle.}.
330:
331: A conspicuous seismic source is seen in the 6~mHz egression power snapshot
332: at 00:42~UT, whose location is indicated by an arrow in all of the frames.
333: A close examination of the source shows that it has two components.
334: By far the most conspicuous component is an intense, compact kernel
335: $\sim$10~Mm located on the penumbral neutral line of the $\delta$-configuration
336: sunspot.
337: Somewhat more diffuse but clearly significant is a secondary, somewhat
338: lenticular signature distributed along the neutral line out to $\sim$15~Mm
339: east and $\sim$30~Mm west of the kernel.
340: These signatures correspond closely with other compact manifestations of
341: the flare.
342: The kernel accounts for approximately 45 per cent of the egression power integrated
343: over the region encompassing the flare signature, with the lenticular
344: component outside of the kernel accounting for the rest.
345:
346: \begin{figure*}
347: \begin{center}
348: %\includegraphics[width=0.78\textwidth, angle=0]{images/timeseries.ps}
349: \includegraphics[width=0.65\textwidth, angle=0]{figure3.eps}
350: \caption{Time series of the 3 and 6 mHz egression and acoustic power (integrated
351: over the neighbourhood of the egression power signatures) are plotted in the top
352: four rows. The dashed vertical lines mark the time of maximum acoustic emission
353: (00:41~UT) at 6~mHz. The relatively extended duration of the acoustic signatures
354: is a result of limits to temporal resolution imposed by truncation of the spectrum
355: (see equation \ref{eq:Deltat}). The bottom two rows show visible continuum irradiance
356: at 1~au from the flaring region along the neutral line in the neighbourhood of the
357: flare. The emission from the neighbourhood of the kernel component of the 6~mHz acoustic
358: source (plot f) is discriminated from the total (plot e).}
359: \label{fig:plots}
360: \end{center}
361: \end{figure*}
362:
363: The 3~mHz egression power snapshots shown in the left column of
364: Fig.~\ref{fig:egpwrsnaps} actually show a considerably stronger seismic emission
365: signature than the 6~mHz signature (right column). But, because of the much greater
366: ambient acoustic emission at this frequency, the 3~mHz signature is not nearly as
367: conspicuous or significant as the 6~mHz signature.
368: It appears to have only a diffuse lenticular component and no conspicuous kernel to
369: match the 6~mHz kernel.
370:
371: It is important to distinguish between the egression power,
372: $|H_+(\textbf{r}, ~t)|^2$, and the local acoustic power,
373: $P(\textbf{r}, ~t)$, which is the square modulus, $|\psi(\textbf{r},
374: ~t)|^2$, of the local wave amplitude $\psi$ at the focus,
375: $\textbf{r}$. Each pixel in a local acoustic power map represents
376: local surface motion as viewed directly from above the photosphere.
377: Each pixel in the egression power map computed by subjacent vantage
378: holography of the surface is a coherent representation of acoustic
379: waves that have emanated downward from the focus, deep beneath the
380: solar surface, and re-emerged into a pupil (see diagram of annulus in
381: Fig.~\ref{fig:egpwrsnaps}a) 15\,--\,45~Mm from the focus.
382: \begin{table*}
383: %\begin{center}
384: \centering
385: \begin{minipage}{120mm}
386: \caption{Energy estimates of the seismic signatures of sunquakes detected
387: to date}
388: \vspace{1em}
389: \renewcommand{\arraystretch}{1.2}
390: \begin{tabular}[thb]{lrcccc}
391: \hline
392: Date & Class & 3~mHz & 6~mHz & 1\,--\,8~\AA ~X-Rays & Visible \\
393: & & (ergs) & (ergs) & (ergs) & (ergs) \\
394: \hline
395: 1996 Jul 09
396: & X2.6
397: & $7.5 \times 10^{27}$
398: & $2.4 \times 10^{26}$
399: & $2.8 \times 10^{29}$
400: & -------------- \\
401: 2001 Sep 09
402: & M9.5
403: & $1.1 \times 10^{27}$
404: & $2.0 \times 10^{26}$
405: & $6.2 \times 10^{28}$
406: & $1.2 \times 10^{30}$ \\
407: 2003 Oct 28
408: & X17.2
409: & $4.7 \times 10^{27}$
410: & $9.4 \times 10^{26}$
411: & $5.0 \times 10^{30}$
412: & -------------- \\
413: 2003 Oct 29
414: & X10.0
415: & $9.4 \times 10^{26}$
416: & $3.5 \times 10^{26}$
417: & $1.5 \times 10^{30}$
418: & $3.8 \times 10^{29}$ \\
419: \textbf{2005 Jan 15}
420: & \textbf{X1.2}
421: & \textbf{$2.4 \times 10^{27}$}
422: & \textbf{$1.0 \times 10^{27}$}
423: & \textbf{$3.4 \times 10^{29}$}
424: & \textbf{$2.0 \times 10^{30}$} \\
425: \hline \\
426: \end{tabular}
427: \end{minipage}
428: \label{tab:table2}
429: %\end{center}
430: \end{table*}
431:
432:
433: Fig.~\ref{fig:acpwrsnaps} shows local acoustic power snapshots of AR10720 at
434: 3~mHz (left column) and 6~mHz (right column) before, during, and after the flare.
435: As in the case of egression power (Fig.~\ref{fig:egpwrsnaps}) all of the local
436: acoustic power maps show a broad acoustic deficit marking the magnetic region.
437: An enhanced local acoustic power halo surrounding the active region is clearly
438: apparent in the 6~mHz snapshots.
439: The acoustic signature of the flare is also clearly visible at 6~mHz.
440: This appears to consist of a pair of kernels, a relatively stronger one nearly
441: coinciding in location with, but slightly east of, the 6~mHz egression power
442: kernel and a weaker one $\sim$10~Mm to west and slight north, lying on the neutral
443: line along which the lenticular component of the 6~mHz egression power is
444: distributed.
445: As in the corresponding egression power snapshot, the 3~mHz local acoustic
446: power snapshots show a stronger but still less conspicuous signature than
447: that at 6~mHz due to a similarly much greater background acoustic power at
448: 3~Mm.
449: \begin{figure*}
450: \begin{center}
451: %\includegraphics[width=0.74\textwidth]{images/wave0.ps}
452: \includegraphics[width=0.6\textwidth]{figure4.eps}
453: \caption{MDI Doppler-difference images showing the expanding ring-shaped wave
454: packet produced by the 2005 January 15 flare.
455: The times shown above respective panels represent one-minute differences.
456: The arrows pointing in the South--East direction (i.e. upper arrows) show
457: the location of the upper-half of the wave front while arrows pointing in
458: the North--West direction (i.e. lower arrows) indicate the lower half of
459: the wave front.
460: Grey-scale at top left expresses Doppler velocity differences in units of
461: ms$^{-1}$ and applies to all frames in the figure.}
462: \label{fig:ripples1}
463: \end{center}
464: \end{figure*}
465:
466: Fig.~\ref{fig:plots} shows plots of the egression and acoustic power time
467: series in the 3 and 6 mHz bands and continuum emission in the neighbourhood
468: of the seismic signature, discriminating continuum emission in the region of the
469: kernel component in the 6~mHz egression power signature from the total.
470: The flare irradiance profiles were extrapolated by applying the assumption that
471: the irradiance is directly proportional to the GONG continuum signature in
472: the neighbourhood of Ni~I~6768~\AA \citep{dl2005}.
473: \begin{figure*}
474: \begin{center}
475: %\includegraphics[width=0.9\textwidth]{images/wave1.ps}
476: \includegraphics[width=0.9\textwidth]{figure5.ps}
477: \caption{MDI Doppler-difference images showing ring-shaped wave packet
478: accelerating outward from the site of the 2005 January 15 flare.
479: The left panel shows the local Doppler signature along the magnetic
480: neutral line during the impulsive phase of the flare at 00:40:30~UT.
481: Arrows in the right panel indicate locations of the ripples
482: propagating outwards from the site of local disturbance 25 minutes later.}%
483: \label{fig:ripples2}
484: \end{center}
485: \end{figure*}
486:
487: The flare of 2005 January 15 produced the most conspicuous acoustic signature
488: of any flare that has produced a detectable seismic emission.
489: This appears to be because such a large fraction of the energy was released
490: into the high-frequency (5\,--\,7~mHz) spectrum, where the competing ambient
491: acoustic power is so far suppressed.
492: Table \ref{tab:table2} shows the energy estimates of the seismic transients radiated
493: into the active region subphotosphere by five flares that have produced
494: conspicuous seismic signatures%
495: \footnote{The energy estimates in Table \ref{tab:table2} were obtained by integrating
496: the egression power over the neighbourhood of the seismic sources (e.g. those shown
497: in Figs.~1g,h for the 2005 January 15 flare).
498: This computation is blind to waves that miss the 15\,--\,45~Mm in the first skip.
499: Comparative seismic holography applied to simulated acoustic transients, and to
500: MDI observations of flares with different sized pupils, indicate that the
501: energies quoted in Table~\ref{tab:table2} account for 80\,--\,95 per cent of the total,
502: depending on the source distribution.}
503: %
504: compared with energy emitted in X-rays in the first 20 minutes of the
505: flare. It should be noted that the 3~mHz energy for the flares
506: preceding the 2005 January 15 flare are actually calculated at 3.5~mHz.
507: Extrapolating through the missing 4\,--\,5~mHz acoustic spectrum for
508: the flare of January 15, we project a total acoustic emission of
509: $\sim 4 \times 10^{20}$~J ($\sim 4 \times 10^{27}$~erg).
510:
511: \subsection{Visible Continuum Emission}
512:
513: Various aspects of visible continuum emission during the 2005 January
514: 15 flare are shown in Figs.~\ref{fig:egpwrsnaps},
515: \ref{fig:acpwrsnaps} and \ref{fig:plots}. The visible-continuum
516: images in Figs.~\ref{fig:egpwrsnaps} and \ref{fig:acpwrsnaps} were
517: obtained by MDI at 00:00~UT, $\sim$37~min before the onset of the
518: flare. We obtained visible continuum maps of AR10720 during the flare
519: from the GONG observatory at Mauna Loa. Technically, the GONG
520: ``continuum intensity maps'' represent a measure of radiation in a
521: $\sim$1~\AA\ bandpass centred on the Ni~I~6768~\AA\ line,
522: whose equivalent width is only a fraction of an \AA\ . Frames c) and
523: d) in Figs.~\ref{fig:egpwrsnaps} and \ref{fig:acpwrsnaps} show the
524: difference in continuum intensity between the GONG images 30 seconds
525: before and after at the time indicated above the frame. Continuum
526: emission is elongated along the magnetic neutral line, corresponding
527: closely to the lenticular component of seismic emission seen at
528: 00:42~UT in Fig.~\ref{fig:egpwrsnaps}h. The brightest emission seen
529: in the intensity difference shown in Fig.~\ref{fig:egpwrsnaps}d
530: comes from a very compact kernel whose location coincides very closely
531: with that of the conspicuous kernel of 6~mHz emission
532: (Fig.~\ref{fig:egpwrsnaps}h).
533:
534: If we assume that the continuum emission emanates isotropically from an opaque surface%
535: \footnote{The assumption is that the specific intensity is independent
536: of the vantage, which implies that the total intensity decreases in
537: proportion to the cosine of the angle of the vantage from normal as a
538: result of foreshortening. If the source was assumed to be optically
539: thin, the resulting energy estimate would be greater by a factor of
540: two. The former appears to be the more realistic estimate for the
541: fraction of visible continuum radiation coming directly from the
542: chromosphere, based on the thesis that ionization of chromospheric
543: hydrogen at the temperature minimum renders the low chromosphere
544: opaque.} the resulting estimate of the total energy emitted in the
545: visible continuum is $2.0 \times 10^{23}$~J ($2.0 \times
546: 10^{30}$~erg). This is $\sim$500 times the total seismic energy we
547: estimate the flare to have emitted into the holographic pupil.
548: Continuum radiation into the neighbourhood of the 6~mHz kernel
549: signature was $6.0 \times 10^{22}$~J ($6.0 \times 10^{29}$~erg). This
550: accounted for $\sim$30 per cent of the total, as compared to 45 per cent of the
551: 6~mHz seismic signature. Continuum emission from in the neighbourhood
552: of the 6~mHz kernel was significantly more sudden than that of the
553: remainder of the acoustic signature.
554:
555: The 2005 January 15 flare contributes to recent findings that relatively
556: small flares can emit disproportionate amounts of acoustic energy
557: \citep{donea2005}. However, even in these cases the fraction of the energy that is released by the flare into the solar interior acoustic spectrum remains relatively small.
558:
559: \subsection{The Surface Ripples}
560:
561: Holography allows us to image the acoustic source of the sunquake when the
562: surface manifestation of the seismic emission is difficult to detect.
563: In the case of the exceptionally powerful seismic transient from the flare
564: of 2005 January 15, the surface signature is quite evident in the raw
565: MDI Doppler observations, a point to which A. Kosovichev (2005, private
566: communication) drew our attention shortly after we reported the discovery
567: of the sunquake to him.
568: To extract the seismic oscillations in the observations we subtracted
569: consecutive MDI Doppler images separated by one minute in time.
570: We applied this Doppler-difference method to a period of observation
571: ($\sim$1 hour) around the time of the flare.
572: Results are shown in Figs.~\ref{fig:ripples1} and \ref{fig:ripples2}.
573:
574: The Doppler signature of the flare is clearly evident at 00:40~UT
575: (Fig.~\ref{fig:ripples2}, left panel). At approximately 20 minutes
576: after the appearance of the flare signature in the sunspot photosphere
577: (at 01:00~UT), we are able to see the seismic response of the
578: photosphere to the energy deposited by the flare in the form of
579: ``ripples'' on the solar surface. In the sequence of one-minute
580: Doppler-difference images in Fig.~\ref{fig:ripples1}, we can see the
581: asymmetrical ring-shaped wave packet propagating from the flare site
582: with the first wave-crest appearing approximately 12\,--\,15~Mm from
583: the flare in a North-Easterly direction. The lower half of the
584: wave-packet has a much smaller amplitude and is propagating in a
585: South-Westerly direction. The arrows in Fig.~\ref{fig:ripples1}
586: indicate the location of the observed wave fronts. The
587: Doppler-difference images in Fig.~\ref{fig:ripples2} show a close-up
588: of the active region at the time of the flare (at 00:40:30~UT, left
589: panel) and the resulting ring-shaped wave packet (at 01:05:00 UT,
590: right panel).
591:
592: The wave-packet was seen to propagate to a maximum distance of
593: approximately 21~Mm from the flare signal, hence travelling a total
594: distance of 6\,--\,9~Mm and lasting for about 8 minutes on the
595: surface, after which the wave amplitude dropped rapidly and the
596: disturbance became submerged in the ambient noise. The lower half of
597: the wave-packet (propagating towards the South\,--\,Western part of
598: the active region, indicated by the lower of the two arrows
599: superimposed on the Doppler-difference images in Figs.~\ref{fig:ripples1} and \ref{fig:ripples2}) was much smaller in
600: amplitude and obscured for much of the 8 minutes.
601: \begin{figure}
602: \begin{center}
603: %\includegraphics[width=0.72\textwidth, angle=0]{images/trace.ps}
604: \includegraphics[width=0.49\textwidth, angle=0]{figure6.eps}
605: \caption{\textit{TRACE} white-light image for AR10720, on 2005 January 15 (00:17:54~UT) with
606: the 12-25 keV \textit{RHESSI} contours (10, 20, 30, 40, 50, 60 and 80 per cent of
607: the maximum flux).
608: The (0,0) coordinates correspond to the location of the seismic source.}%
609: \label{fig:trace}
610: \end{center}
611: \end{figure}
612:
613: \begin{figure*}
614: \begin{center}
615: %\includegraphics[width=0.72\textwidth, angle=0]{images/hxr3.ps}
616: \includegraphics[width=0.7\textwidth, angle=0]{figure7.ps}
617: \caption{The HXR flux in the 4\,--\,25 (0.5\,--\,4~\AA) and 25\,--\,50~keV energy ranges as observed by \textit{GOES}
618: (dot-dashed curve, multiplied by a factor of $10^{6.5}$~Wm$^{-2}$) and \textit{RHESSI} (bold curve,
619: multiplied by a factor of $10^{-2.8}$~counts) respectively. The solid curve represents
620: the 1-minute mean averages of the Doppler velocity (ms$^{-1}$) in the quake region for
621: the period 00:00:00\,--\,01:20:00~UT.
622: The dashed vertical line represents the observed maximum emission at 6~mHz.}
623: \label{fig:Xrays}
624: \end{center}
625: \end{figure*}
626:
627: \subsection {Hard X-Ray Emission}
628:
629: The \textit{TRACE} data for the 2005 January 15 flare in the white-light channel
630: have a variable cadence for the period 00:00:00\,--\,01:00:00~UT.
631: Fig.~\ref{fig:trace} shows the \textit{TRACE} white-light image taken at
632: 00:17:54~UT, approximately 10~minutes before the onset of the X1.2
633: flare with the \textit{RHESSI} 12-25~keV contours overlaid. The \textit{RHESSI} HXR
634: image is averaged over the period 00:41:33\,--\,00:42:34~UT. The time
635: of peak intensity in this energy band occurs at 00:42:04~UT, a close
636: temporal correlation with the maximum of the seismic emission detected
637: at 6~mHz. The HXR emission is thought to represent bremsstrahlung
638: emission from high-energy coronal electrons impinging into the
639: chromosphere \citep{b1971}.
640:
641: The 12\,--\,25~keV emission at 00:42:00~UT extends along the neutral
642: magnetic line. We identify three compact HXR sources (see the numbers
643: in Fig.~\ref{fig:trace}). Source 2 is the strongest, while source 3
644: is the weakest. These could represent the foot-points of a complex
645: magnetic loop. However, source 1 (which emits 50 per cent of the total flux)
646: spatially coincides with the lenticular component of the 6~mHz seismic
647: source (see Fig.~\ref{fig:egpwrsnaps}). This reinforces the role of
648: non-thermal particles in supplying the energy that drives the seismic
649: emission. Similar comparisons have been observed in other flares \citep{dl2005}.
650:
651: Furthermore, Fig.~\ref{fig:Xrays} reveals that the velocity impulse
652: of the flare in the sunspot photosphere was almost as sharp as the HXR
653: flux detected in the 4\,--\,25~keV (0.5\,--\,4~\AA) energy range by
654: the \textit{GOES} satellite, but the maximum HXR emission (observed at
655: $\sim$00:43:00~UT) appears to have occurred $\sim$2 minutes after the
656: maximum velocity depression at the photosphere (00:41:00 UT). In
657: fact, a sudden drop of approximately 100~ms$^{-1}$ in the mean
658: velocity of the Doppler signal (an upflow) is observed in the 3 minute
659: period from 00:38:00-00:41:00~UT. The \textit{RHESSI} HXR peak in the higher
660: energy band of 25\,--\,50~keV plotted in Fig.~\ref{fig:Xrays} occurs
661: at $\sim$00:41:00~UT, which temporally coincides with both the maximum
662: of the seismic source at 6~mHz and the velocity depression at the
663: photosphere. We also note that the peak emission in the 3\,--\,12~keV
664: energy band detected by both \textit{GOES} (1\,--\,8~\AA) and \textit{RHESSI} (occurring
665: at $\sim$00:44:00 and $\sim$00:47:00~UT respectively, but not
666: plotted), also have a close temporal correlation with the maximum of
667: the seismic emission.
668:
669:
670: \section{Discussion and Summary}
671:
672: The X1.2-class flare of 2005 January 15 produced one of the most
673: powerful sunquakes detected to date and by far the most conspicuous,
674: on account of exceptionally powerful emission above 5~mHz from a
675: compact source. Certain qualities exhibited by the flare of January
676: 15 are shared by all other known acoustically active flares. The
677: first is the coincidence between strong compact acoustic sources and
678: nearby signatures of HXR emission. This suggests that high-energy
679: particles supply the energy that drives the acoustic emission, and it
680: is evident from the electromagnetic emission attributed to these
681: particles that they contain more than sufficient energy for this
682: purpose. The appearance of sudden, conspicuous white-light emission
683: from the flare of 2005 January 15 closely co-spatial with the location and
684: morphology of the holographic signatures is similarly characteristic
685: of all other known acoustically active flares so far.
686:
687: \citet{kz1998} proposed that seismic emission into the solar interior
688: in sunquakes is the continuation of a chromospheric shock and
689: condensation resulting from explosive ablation of the chromosphere and
690: propagating downward through the photosphere into the underlying solar
691: interior. Chromospheric shocks are well known under such
692: circumstances, based on red-shifted H$\alpha$ emission at the flare
693: site at the onset of the flare. The theory of their dynamics was
694: worked out at length by \citet{fisher1985a,fisher1985b,fisher1985c} and
695: others since. The hypothesis that solar interior emission is a direct
696: continuation of such shocks was considered by \citet{dl2005}, who found
697: the signature of a strong downward-propagating chromospheric transient
698: in Na~D$_1$-line observations of the flare of 2003 October 29.
699: However, we are now aware of similar chromospheric transient
700: signatures with no significant attendant holographic signature to
701: indicate seismic emission into the active region
702: subphotosphere\footnote{An example is seen in the western foot-point of
703: the magnetic loop that hosted 2003 October 29 flare (see right frame
704: in second row of fig.~8 labelled ``Red \hbox{[0 min]}'' in
705: \citet{dl2005}). The corresponding signature of sudden white-light
706: emission, seen in fig.~9 of the same, shows only a weak signature at
707: the same location. The seismic signature, seen in the lower left
708: corner of the upper right frame of fig.~11 of the same, shows
709: correspondingly weak seismic emission.}%
710: %
711:
712: In these instances, the signature of sudden white-light emission is
713: relatively weak. Following \citet{machado1989}, \citet{donea2005}
714: proposed to attribute the lack of seismic emission where there is a
715: strong chromospheric transient but only a weak or absent white-light
716: signature to strong radiative damping that depletes the chromospheric
717: transient before its arrival into the low photosphere.
718:
719: In all acoustically active flares encountered to date, there is a
720: strong spatial correlation between the sources of seismic emission and
721: sudden white-light emission. This remains conspicuously the case for
722: the flare of 2005 January 15, as a comparison between Figs.~1d and 1h
723: shows. In some instances, e.g. the large flares of 2003 October 29,
724: the source of the white-light emission has been much more extensive
725: than the source of the acoustic emission, the former many times the
726: area of the latter and encompassing it. However, in these instances
727: the temporal profile of visible continuum emission significantly away
728: from any of the sites of seismic emission has been comparatively
729: sluggish and diffuse. What has particularly and consistently
730: distinguished the white-light signature in the neighbourhood of the
731: acoustic emission has been the suddenness of its appearance, on a time
732: scale of a minute or two, and possibly considerably less than a minute
733: given that the observations of continuum emission associated with
734: flares to date have been limited to a cadence of one minute.
735:
736: It should be kept in mind that the energies released in known seismic
737: transients have invariably been a small fraction of the energy
738: released into the visible continuum spectrum. The actual fraction has
739: varied considerably, from a few millionths, in the case of the flare
740: of 2003 October 29 \citep{dl2005}, to a few thousandths, for the flare
741: of 2005 January 15. However, if only the sudden-onset continuum
742: emission in the neighbourhood of the seismic sources is included, then
743: the ratio for the flare of 2003 October 29 is similar to that of the
744: flare of 2005 January 15. This is what is listed in
745: Table~\ref{tab:table2} of this study.
746:
747: {\it The close coincidence between the locations of sudden white-light
748: emission and seismic transient emission in all acoustically active
749: flares to date suggests that a substantial component of the seismic
750: emission seen is a result of sudden heating of the low photosphere
751: associated with the visible continuum emission seen.} A complete
752: analysis of wave emission as a result of transient heating involves
753: detailed considerations of energy and momentum balance. An
754: approximate account of these was undertaken by \citet{donea2005}.
755: Basic considerations of momentum balance are described in Section 4.3
756: of \citet{donea2005}, adapting the discussion by
757: \citet{canfield1990} of momentum balance in chromospheric
758: transients to transients similarly excited by sudden heating in the
759: low photosphere.
760:
761: \citet{donea2005} devised a rough, preliminary physical model to
762: estimate the energy of the seismic transient to be emitted as a result
763: of sudden, momentary heating of the low photosphere to a degree
764: consistent with the transient white-light signature closely coincident
765: with the seismic source in the M9.5-class flare of 2001 September 9.
766: Their estimate expressed the energy, $E$, of the seismic transient in
767: terms of the thermal energy, $U$, radiated or dissipated into the low
768: photosphere, and the fractional increment, $\delta p/p$, in pressure
769: that would result from the heating:
770: \begin{equation}
771: E ~=~ {1 \over 2} H {(\delta p)^2 \over p}
772: ~=~ {1 \over 3} ~{\delta p \over p} ~\delta U,
773: \label{eq:EdUdp}
774: \end{equation}
775: where $H$ is the $e$-folding height of the photospheric density.
776: This relation appears to be roughly consistent with the few-percent continuum
777: intensity variations observed for the flare of 2001 September 9, if the relation
778: between $\delta p$ and $\delta I$, the variation in continuum intensity, can
779: be approximated by the Stefan-Boltzmann law,
780: \begin{equation}
781: {\delta p \over p} ~=~ {\delta T \over T}
782: ~=~ {1 \over 4}{\delta I \over I},
783: \label{eq:boltzmann}
784: \end{equation}
785: and the heating is accomplished within a duration not excessively
786: longer than $\tau _{ac} = 1/\omega _{ac} \sim 40$~s, where
787: $\omega_{ac}$ is the acoustic cutoff frequency in the low photosphere
788: (see Section 4 of \citet{dl2005}). A similar exercise applied to
789: the flare of 2005 January 15 leads to similar results. In fact, the
790: ratio of the seismic energy to the electromagnetic energy is roughly
791: the same for both of these flares, as are the mean intensity
792: increments if credible boundaries are chosen over which to take the
793: mean. To the extent that we can resolve the fine details, acoustic
794: emission from the flare of 2005 January 15 could reasonably be the
795: result of photospheric heating similar to that of the 2001 September 9
796: flare but over approximately twice the area. Differences between the
797: two flares could be attributed to differing photospheric or
798: subphotospheric thermal conditions and differing magnetic fields, for
799: which the foregoing approximation contains no account.
800:
801: A detailed examination of the physics of heated magnetic photospheres
802: is needed to lend credibility to the hypothesis that seismic emission
803: from acoustically active flares is driven by sudden heating of the low
804: photosphere by any mechanism whatever. At this point we will only say
805: that this hypothesis appears to be consistent with our present limited
806: understanding of the observations. However, there is some controversy
807: as to the implications of visible continuum emission during flares
808: with respect to heating of the low photosphere. In the case of the
809: flares of 2003 October 28\,--\,29, the signature of high-energy protons
810: along with the particles that gave rise to X-ray emission lent
811: considerable weight to the interpretation of visible continuum
812: emission in terms of a heated low photosphere, as protons are
813: sufficiently massive to penetrate to the bottom of the photosphere and
814: heat it directly by collisions. The flare of 2005 January 15, on the
815: other hand, confronts us with an instance of intense seismic emission
816: with no indication of high-energy protons among the energetic
817: particles that supply the energy on which the acoustic emission
818: depends. Energetic electrons consistent with HXR signatures cannot
819: penetrate into the low photosphere in anywhere near sufficient numbers
820: to account for the heating required by the seismic signatures
821: \citep{metcalf1990}. \citet{chen05} also affirm that the white-light flare
822: signatures highlight the importance of radiative back-warming in transporting
823: the energy to the low photosphere when direct heating by beam electrons is impossible.
824:
825: In such cases, it appears to be well established that the origin of
826: white-light emission would have to be entirely in the chromosphere,
827: where energetic electrons dissipate their energy
828: \citep{metcalf1990,zharkova1991,zharkova1993}, mainly by ionizing
829: previously neutral chromospheric hydrogen approximately to the depth
830: of the temperature minimum. Nevertheless, even in these instances, it
831: appears that the low photosphere itself would be significantly heated
832: as a secondary, but more or less immediate, effect of chromospheric
833: ionization. This is primarily the result of Balmer and Paschen
834: continuum edge recombination radiation from the overlying ionized
835: chromospheric medium, approximately half of which we assume radiates
836: downward and into the underlying photosphere. When the intensity,
837: $\delta I$, of this downward flux is commensurate with a temperature
838: perturbation, $\delta T$, consistent with the Stefan-Boltzmann law
839: (equation [\ref{eq:boltzmann}]), the result of such a flux is
840: understood to be heating of the low photosphere such as to bring about
841: a temperature increment of roughly this order within a few seconds
842: \citep{donea2005,machado1989, metcalf2003}. Heating of the photosphere
843: by the mechanism described above is known as ``back-warming''
844: \citep{metcalf2003} and a substantial fraction of the continuum
845: emission seen in white-light flares is thought to represent the
846: downward flux from an ionized chromosphere thermally re-emitted by the
847: now heated photosphere. In this light, the strong correlation between
848: sources of white-light and seismic emission into the solar interior
849: might be regarded as strong support for the back-warming hypothesis
850: when this relation persists in flares devoid of protons among the
851: high-energy particles that drive the flare. This is certainly the
852: case for the flare of 2005 January 15.
853:
854: \citet{dl2005} and \citet{donea2005} summarize our understanding of the relationship
855: between the efficiency of seismic emission and the suddenness of the heating
856: that drives the seismic transient.
857: Based on these considerations, one has to suspect that the perceptibly more sudden
858: profile of continuum emission in the neighbourhood of the kernel component of the
859: 6~mHz emission accounts to a significant degree for the disproportionate power in
860: the 6~mHz egression-power signature.
861: This is one of the many aspects of flare acoustics that would benefit from detailed
862: modelling, including a careful account of magnetic forces.
863:
864: At this point, our understanding of seismic emission from flares remains
865: relatively superficial.
866: However, evidence for the general involvement of photospheric heating is
867: now considerable.
868: What is needed for further understanding is detailed modelling with a careful
869: account of the physics, including radiative transfer and magnetic forces
870: in realistic sunspot photospheres and subphotospheres.
871: With such an understanding, acoustic emission from flares could contribute major
872: benefits to seismic diagnostics of active region subphotospheres and the physics
873: of mode conversion \citep{cally2000}.
874:
875: %A clear understanding of the physics of seismic emission could help us to penetrate
876: %major technical issues in active region seismology, such as the effects of molecular
877: %contamination of seismic signatures in sunspots.
878: %In the non-flaring sunspot photosphere, contamination by molecular lines may be
879: %negligible, simply because apparent shifts in the wavelength of the line used for
880: %Doppler measurements are constant and therefore do not find their way into the
881: %acoustic spectrum.
882: %However, visible and UV radiation produced by white-light flares is probably
883: %capable of disassembling such molecules, opening the likelihood of false Doppler
884: %transients under conditions that would give rise to significant seismic emission.
885: %
886: %Lines of molecules such as TiO abound in sunspot umbrae.
887: %Because both titanium and oxygen are far less abundant than hydrogen in the solar
888: %photosphere, it may take this molecule minutes or hours to recover from a white
889: %light flare.
890: %It would be most interesting to see what sort of perturbations a white-light
891: %flare would introduce into an optically thin TiO line and what the time scale
892: %of its recovery would be.
893:
894: \section*{Acknowledgments}
895:
896: We have benefited greatly from the insights of Drs. T. Metcalf and V. Zharkova.
897: We also greatly appreciate comments by Dr. A. McClymont.
898: Dr. A. Kosovichev drew our attention to certain interesting aspects of the
899: surface ripples caused by the seismic transient, of which the holographic
900: signatures represent the sources rendered in Figs.~1g,h.
901: The research reported here heavily utilized data obtained by the MDI instrument
902: on the \textit{SOHO} spacecraft, operated by the National Aeronautics and Space
903: Administration (NASA) and the European Space Agency (ESA).
904: This work further utilized data obtained by the GONG++ project, managed by the
905: National Solar Observatory, a Division of the National Optical Astronomy
906: Observatories, which is operated by AURA, Inc. under a cooperative agreement
907: with the National Science Foundation.
908: This research was supported by grants and contracts from the Astronomy and
909: Stellar Astrophysics Branch of the National Science Foundation.
910: It was also supported by NASA's Sun-Earth Connection/Solar Heliospheric Physics
911: Program.
912:
913:
914: \begin{thebibliography}{}
915:
916: \bibitem[\protect\citeauthoryear{Braun}{1995}]{braun1995}
917: Braun D. C., 1995, ApJ, 451, 859
918:
919: \bibitem[\protect\citeauthoryear{Braun, Duvall \& LaBonte}{1988}]{bdl1988}
920: Braun D. C., Duvall T. L. Jr., LaBonte B. J., 1988, ApJ, 335, 1015
921:
922: \bibitem[\protect\citeauthoryear{Braun et al.}{1998}]{blff1998}
923: Braun D. C., Lindsey C., Fan Y., Fagan M., 1998, ApJ, 502, 968
924:
925: \bibitem[\protect\citeauthoryear{Braun \& Lindsey}{1999a}]{bl1999a}
926: Braun D. C., Lindsey C., 1999a, ApJ, 510, 494
927:
928: \bibitem[\protect\citeauthoryear{Braun \& Lindsey}{1999b}]{bl1999b}
929: Braun D. C., Lindsey C., 1999b, ApJ, 513, L79
930:
931: \bibitem[\protect\citeauthoryear{Brown}{1971}]{b1971}
932: Brown J. C., 1971, Solar Phys., 18, 489
933:
934: \bibitem[\protect\citeauthoryear{Cally}{2000}]{cally2000}
935: Cally P. S., 2000, Solar Phys., 192, 395
936:
937: \bibitem[\protect\citeauthoryear{Canfield et al.}{1990}]{canfield1990}
938: Canfield R. C., Zarro D. M., Metcalf T. R., Lemen J. R., 1990, Solar Phys., 348, 333
939:
940: \bibitem[\protect\citeauthoryear{Chen \& Ding}{2006}]{chen05}
941: Chen Q. R., Ding M. D., 2006, ApJ , 641, 1217
942:
943: \bibitem[\protect\citeauthoryear{DeForest}{2004}]{deforest2004}
944: DeForest C. E., 2004, Solar Phys., 219, 3
945:
946: \bibitem[\protect\citeauthoryear{Donea, Braun \& Lindsey}{1999}]{dbl1999}
947: Donea A. -C., Braun D. C., Lindsey C., 1999, ApJ, 513, L143
948:
949: \bibitem[\protect\citeauthoryear{Donea \& Lindsey}{2005}]{dl2005}
950: Donea A. -C., Lindsey C., 2005, ApJ, 630, 1168
951:
952: \bibitem[\protect\citeauthoryear{Donea et al.}{2005}]{betal2006}
953: Donea A. -C., Besliu D., Cally P. S., Lindsey C., 2005, in Leibacher J.,
954: Uitenbroek H., Stein B., eds, ASP Conf. Ser., Solar MHD: Theory and Observations - A
955: High Spatial Resolution Perspective, Astron. Soc. Pac., in press
956:
957: \bibitem[\protect\citeauthoryear{Donea et al.}{2006}]{donea2005}
958: Donea A. -C., Besliu-Ionescu D., Cally P. S., Lindsey C., Zharkova V. V., 2006, Solar Phys., accepted June 2006
959:
960: %\bibitem[\protect\citeauthoryear{Donea et al.}{in preparation}]{donea2006}
961: %Donea A. C., Martinez Oliveros J. C., Cally P. S., Moradi H., 2006, in preparation
962:
963: \bibitem[\protect\citeauthoryear{Fisher, Canfield \& McClymont}{1985a}]{fisher1985a}
964: Fisher G. H., Canfield R. C., McClymont A. N., 1985a, ApJ, 289, 414
965:
966: \bibitem[\protect\citeauthoryear{Fisher, Canfield \& McClymont}{1985b}]{fisher1985b}
967: Fisher G. H., Canfield R. C., McClymont A. N., 1985b, ApJ, 289, 425
968:
969: \bibitem[\protect\citeauthoryear{Fisher, Canfield \& McClymont}{1985c}]{fisher1985c}
970: Fisher G. H., Canfield R. C., McClymont A. N., 1985c, ApJ, 289, 434
971:
972: \bibitem[\protect\citeauthoryear{Hudson}{1972}]{hudson1972}
973: Hudson H. S., 1972, Solar Phys., 24, 414
974:
975: \bibitem[\protect\citeauthoryear{Kosovichev \& Zharkova}{1998}]{kz1998}
976: Kosovichev A. G., Zharkova V. V., 1998, Nature, 393, 317
977:
978: \bibitem[\protect\citeauthoryear{Lindsey and Braun}{1999}]{lb1999}
979: Lindsey C., Braun D. C., 1999, ApJ, 510, 494
980:
981: \bibitem[\protect\citeauthoryear{Lindsey and Braun}{2000}]{lb2000}
982: Lindsey C., Braun D. C., 2000, Solar Phys., 192, 261
983:
984: \bibitem[\protect\citeauthoryear{Machado et al.}{1989}]{machado1989}
985: Machado M. E., Emslie A. G., Avrett E. H., 1989, Solar Phys., 124, 303
986:
987: \bibitem[\protect\citeauthoryear{Metcalf, Canfield \& Saba}{1990}]{metcalf1990}
988: Metcalf T. R., Canfield R. C., Saba J., 1990, ApJ, 365, 391
989:
990: \bibitem[\protect\citeauthoryear{Metcalf et al.}{2003}]{metcalf2003}
991: Metcalf T. R., Alexander D., Hudson H. S., Longcope D., 2003, ApJ, 595, 483
992:
993: \bibitem[\protect\citeauthoryear{Scherrer et al.}{1995}]{setal1995}
994: Scherrer P. H. et al., 1995, Solar Phys., 162, 129
995:
996: \bibitem[\protect\citeauthoryear{Zharkova \& Kobylinskii}{1991}]{zharkova1991}
997: Zharkova V. V., Kobylinskii V. A., 1991, Sov. Astron. Lett., 17, 34
998:
999: \bibitem[\protect\citeauthoryear{Zharkova \& Kobylinskii}{1993}]{zharkova1993}
1000: Zharkova V. V., Kobylinskii V. A., 1993, Solar Phys., 143, 249
1001:
1002: \end{thebibliography}
1003:
1004: \end{document}
1005: