1:
2: \documentclass[12pt, preprint]{aastex}
3:
4: \shorttitle{Deep Meridional Circulation}
5: \shortauthors{Braun and Birch}
6:
7: \begin{document}
8:
9: \title{Prospects for the Detection of the Deep Solar Meridional Circulation}
10:
11: \author{D.~C.\ Braun, A.~C.\ Birch}
12: \affil{NorthWest Research Associates, CoRA Division, 3380 Mitchell Lane,
13: Boulder, CO 80301, USA}
14: \email{dbraun@cora.nwra.com aaronb@cora.nwra.com}
15:
16: \begin{abstract}
17: We perform helioseismic holography to assess the noise
18: in p-mode travel-time shifts which would form the basis of
19: inferences of large-scale flows throughout the
20: solar convection zone.
21: We also derive the expected travel times from a parameterized
22: return (equatorward) flow component of the meridional circulation
23: at the base of the convection zone
24: from forward models under the assumption of
25: the ray and Born approximations.
26: From estimates of the signal-to-noise ratio
27: for measurements focused near the base of
28: the convection zone, we conclude that the helioseismic detection of
29: the deep meridional flow including the return component
30: may not be possible using data spanning an interval less
31: than a solar cycle.
32:
33: \end{abstract}
34:
35: \keywords{Sun: helioseismology, interior}
36:
37: \section{Introduction}
38:
39: Among all known large-scale flows in the Sun, the
40: meridional circulation has particular significance
41: because of its role in the transport of angular momentum and
42: magnetic flux across a wide range of latitudes within the
43: convection zone. Consequently, it is a significant component of
44: models of the dynamics
45: of rotating stellar convection zones, dynamos, and the solar cycle
46: \citep{Glatzmaier1982, Choudhuri1995, Dikpati2001, Wang1991,
47: Wang2002, Hathaway2003, Hathaway2004, Dikpati2006, Dikpati2007}.
48:
49: Measurements of the surface manifestation of meridional
50: circulation have typically indicated poleward flows
51: between 10 and 20 m s$^{-1}$ \citep[e.g.][]{Hathaway1996}.
52: Although frequencies of global p modes are insensitive (to
53: first order) to the meridional circulation, the flows have been
54: detected with a variety of local seismic methods
55: \citep[e.g.][]{Giles1997, Giles1998, Braun1998, Gonzalez1999,
56: Giles2000, Haber2002, Hughes2003, Zhao2004,
57: Chou2005, Gonzalez2006, Svanda2007, Mitra-Kraev2007, Gonzalez2008}.
58: Many of these studies have focused their attention on
59: the meridional circulation near the surface (e.g.\ within a few
60: tens of Mm below the surface), and only a few attempts
61: have been made to deduce the properties of the deeper
62: components. Among the most comprehensive analyses is the work
63: of \citet{Giles2000} which is based on models of time-distance measurements
64: using over two years of Michelson Doppler Imager
65: \citep[MDI;][]{Scherrer1995}
66: Dopplergrams. These models included two general
67: solutions for the meridional circulation as a function of depth
68: and latitude: the first (hereafter ``Giles' model A'') without any
69: constraint on mass conservations, and the second (Giles' model B)
70: with an imposed mass conservation. Only Giles' model B exhibited a return
71: flow while the other showed exclusively poleward flows throughout
72: the convection zone. Each model was consistent with the travel-time
73: measurements within their range of errors \citep{Giles2000}.
74:
75: In the frequency--wavenumber range of p modes
76: propagating through the bottom half of the convection zone,
77: the random noise present in most current helioseismic measurements
78: is dominated by realization noise caused by stochastic excitation of the
79: p modes near the solar surface. For the exploration of large-scale
80: flows such as meridional circulation this can be reduced by observing
81: more of the Sun (e.g.\ the far side of the Sun which is not currently
82: accessible to helioseismic instruments) or by employing datasets
83: with longer temporal duration.
84: With over a decade of helioseismic observations from both
85: the Global Oscillations Network
86: Group (GONG; \url{http:/gong.nso.edu/data}) and MDI (\url{http:/soi.stanford.edu/data}) now available it is worthwhile to revisit the issue of
87: the deep meridional circulation.
88: In this paper, we explore the prospects
89: for helioseismic detection of the return component of the meridional
90: circulation in the deep
91: solar convection zone by applying helioseismic holography
92: to MDI observations to assess the random noise
93: in travel-time shifts which would form the basis for the
94: inference of large-scale flows. Our analysis and resulting
95: noise estimates are described in
96: \S~\ref{noise}. We estimate the expected signal from
97: a plausible return component of meridional circulation using forward modeling procedures
98: described in \S~\ref{models}.
99: This is followed in \S~\ref{discussion} by a discussion of
100: the implications of these results.
101:
102: \section{Noise Assessment} \label{noise}
103:
104: Helioseismic holography (hereafter HH) is a method which
105: computationally extrapolates the surface acoustic field
106: into the solar interior \citep{Lindsey1997, Lindsey2000}
107: in order to estimate the amplitudes of the waves propagating into
108: and out of a focus point at a chosen depth and position in the solar
109: interior. The magnitudes and phases of these amplitudes,
110: called the ingression and egression, are used to detect
111: flows and other perturbations to the waves.
112: The method employed for horizontal flow diagnostics is based on the
113: egressions and ingressions computed in the {\it lateral
114: vantage} employing pupils spanning 4 quadrants
115: extending in the east, west, north and south directions
116: from the focus \citep{Braun2004, Lindsey2004, Braun2007}.
117: In the lateral vantage, the p modes sampled by the pupil
118: propagate through the focal point in directions inclined up
119: to $\pm 45^\circ$ from the direction parallel to the
120: surface \citep[see Figure 3 of][]{Braun2007}.
121: A difference in the travel times between
122: waves traveling from one pupil to its opposite and
123: waves traveling in the reverse direction
124: is produced by flows along the path of the waves.
125: In particular, the travel time differences,
126: $\delta{\tau}_{\rm ew}$ and $\delta{\tau}_{\rm ns}$
127: derived from the east--west
128: and north--south quadrant pairs, respectively,
129: provide the HH signatures sensitive to the two components of
130: the horizontal flow.
131: The sign of the travel-time difference is such that a
132: northward velocity component will produce a negative
133: value of $\delta{\tau}_{\rm ns}$.
134: The lateral-vantage geometry samples
135: more than 70\% of the wave modes which pass through
136: the focus. The remaining waves, propagating more vertically than
137: the waves appearing in the pupil, are substantially less sensitive to
138: horizontal flows near the focus.
139: Table~\ref{tbl-1} lists the focus depths and the pupil
140: radii used in this study. The pupil radii are
141: defined from ray theory. The range
142: of (spherical-harmonic) mode degrees ($\ell$) at
143: 4 mHz, selected
144: by each pupil, is also listed in the table. The lower
145: $\ell$ value denotes the modes propagating at
146: $\pm 45^\circ$
147: from the horizontal direction which propagate through the
148: focus and reach the surface at either the inner or
149: outer pupil radius. The highest $\ell$ value listed
150: indicates modes propagating horizontally through the focus.
151: The focus depths extend down to the base of the convection zone.
152: However, the analysis is conceptually similar to previous near-surface
153: measurements \citep{Braun2007}.
154:
155: Three weeks of full disk Dopplergrams with one minute
156: cadence, obtained from MDI
157: were used in this study.
158: The data set spans the interval from 1996 June 25 to July 16,
159: and coincides with a period of very low magnetic activity
160: on the Sun. Smaller spans of data at other epochs (2002 March
161: and 2003 October) were also
162: examined. Travel-time maps made at all three epochs
163: exhibit similar noise characteristics, and
164: we show here only the results using the 1996 data.
165: The following steps summarize the general data reduction:
166: 1) a projection of each 24 hr segment of full-disk data onto
167: nine Postel projections (each extending $180^\circ \times 180^\circ$)
168: centered on grid points separated by $40^\circ$ in
169: heliographic latitude and
170: central meridian distance referenced to midday
171: and rotating with the Carrington rate,
172: 2) temporal detrending by subtraction of a linear
173: fit to each pixel signal in time, 3) removal of poor
174: quality images, identified by a five-$\sigma$ deviation of
175: any pixel from the linear trend , 4) Fourier transform of
176: the data in time,
177: 5) extraction of the frequency
178: bandpass spanning 2.5 to 5.5 mHz, 6) computation of Green's
179: functions over the appropriate pupil, 7) computation
180: of ingression and egression amplitudes by a 2D convolution
181: of the data with the Green's functions,
182: 8) computation of the travel-time difference maps, and
183: 9) extraction and remapping of the central $40^\circ \times 40^\circ$ portion of
184: each region to form mosaics in heliographic coordinates
185: spanning $120^\circ \times 120^\circ$.
186: The Green's functions (step 6) were computed using
187: the eikonal approximation \citep[][]{Lindsey1997, Lindsey2000} in
188: spherical coordinates. The large size of the Postel's projections
189: is dictated by the large pupil required for the deepest focii
190: in Table~\ref{tbl-1}.
191: The 2D convolution (step 7) is a time-saving convenience,
192: appropriate for the type of preliminary noise estimates we are
193: interested in, but distorts the resulting travel-time difference maps.
194: This occurs because of a geometrical
195: mismatch between the fixed annular pupil assumed for the
196: convolution operation and the correct
197: pupil whose shape varies with position in the Postel projection.
198: This deviation worsens as the horizontal position of the focus is moved
199: away from the central (tangent) point of the Postel projection,
200: and the effects of this can therefore be constrained
201: by combining maps made using multiple locations of
202: the tangent point. The consequences of this distortion
203: for the results presented here are discussed below.
204:
205: Figure~\ref{maps} shows selected maps of the mean
206: $\delta{\tau}_{\rm ns}$ and standard deviation $\sigma_0$ of the north-south
207: travel-time difference
208: over twenty consecutive (24-hr duration)
209: sets of measurements. Each map covers an area
210: of $120^\circ$ in central meridian distance (CMD) and
211: heliographic latitude (B).
212: The maps for one day of data (June 27) were
213: not included in further analysis due to an anomalously high amount of
214: poor images. Near the surface (e.g.\ Figure~\ref{maps}a),
215: the meridional flow produces a distinct
216: negative (positive) travel-time difference in the north (south) hemisphere.
217: As the focus depth increases, this signature becomes less visible.
218: A distinct pattern near the poles is also evident and increases
219: significantly with greater focus depth. This pattern is opposite
220: in sign of the meridional signature and is clearly an artifact
221: centered on the position of disk center (about $+3^\circ$)
222: as observed by MDI.
223:
224: Remarkably, the maps of the standard deviation (Figs~\ref{maps}d -
225: \ref{maps}f) indicate that, apart from the vicinity of the solar limb,
226: the noise for a single travel-time measurement is fairly
227: constant ($\sigma_0 \approx 4 \rm{sec}$) with focus depth.
228: There is a granularity in these
229: maps which becomes courser with depth and is related to
230: the increase in the horizontal wavelength of the modes
231: used to make the measurements. An increase in the
232: standard deviation near the solar limb is evident.
233: In addition, there is a noticeable
234: excess of $\sigma_0$ near positions
235: (CMD, B) = $(\pm 20^\circ, \pm 20^\circ)$ which are
236: the corners of the individual subregions of
237: the mosaic. The noise in these corners is about 15\%
238: above the values near the Postel tangent points for all
239: focus depths used here. This feature likely
240: results from the use of the 2D convolution of Postel's
241: projections described earlier.
242:
243: Due to signal-to-noise issues we assume any current or future attempt to deduce
244: properties of the deep meridional circulation will
245: make use of longitudinal averaging and very likely also involve at least
246: modest smoothing in latitude. Figure~\ref{noise_vs_depth}
247: shows the standard deviation ($\sigma_{\rm a}$) of averages of $\delta \tau_{\rm ns}$
248: over strips spanning $120^\circ$ in CMD and $15^\circ$
249: in B. Unlike the standard deviation ($\sigma_0$) corresponding to specific
250: horizontal focus positions which show little or no variation with
251: focus depth, the standard deviation of the mean ($\sigma_{\rm a}$)
252: increases with depth at all latitudes. This
253: is a consequence of having fewer independent measurements within a fixed
254: area as the depth increases and is analogous to the common problem in global
255: helioseismology of having fewer modes in which to deduce either
256: structural perturbations or flows at greater depths within the Sun.
257: There is also an increase in noise for measurements at high latitudes,
258: consistent with the maps of $\sigma_0$ (Figure~\ref{maps}).
259: Significantly, the values of $\sigma_{\rm a}$ for the $15 - 30^\circ$
260: strips are very close to the results for the $0 - 15^\circ$ strip.
261: This offers some
262: assurance that the contribution of noise due to the 2D convolution
263: (which should preferentially influence the
264: $15 - 30^\circ$ strip)
265: does not add substantially to the results shown in
266: Figure~\ref{noise_vs_depth}.
267:
268:
269: \section{Forward Models} \label{models}
270:
271: We use both the Born and ray approximations to estimate the travel-time
272: shifts that would be caused by a return flow near the base of the convection
273: zone.
274: For a discussion of the ray approximation see \citet{Giles2000}.
275: We used the numerical approach of \citet{JCD89} for computing
276: ray paths in spherical geometry.
277:
278: For this letter we also make rough estimates of the
279: sensitivity of HH travel times
280: to weak, steady, and horizontally uniform flows by approximating
281: the convection zone as a plane-parallel layer.
282: The functions which describe the linear
283: sensitivity of the power spectrum to a horizontally uniform flow
284: can be computed using a generalization of the
285: Born-approximation based approach of \citet{Gizon2002} and \citet{Birch2007}.
286: We used the normal-mode summation Green's functions from \citet{Birch2004},
287: though with the eigenfunctions for a spherical Sun in place of those for a
288: plane-parallel version of model S.
289: We used the source model of \cite{Birch2004}.
290: Changes in the the power spectrum may easily be related to changes in the
291: ingression-egression correlation through the expression for
292: the expectation value of the correlation. The result is a set of
293: sensitivity kernels which relate the correlations (and thus the
294: travel-time shifts) with the flows.
295:
296: Using the sensitivity functions we estimate the travel-time shifts caused
297: by deep return flows of the form $v(z) = A \cos\{\pi(r-r_{\rm c})/\Delta r\}$
298: for $r_{\rm c}<r<r_{\rm c}+\Delta r$ and $v(z)=0$ otherwise.
299: Here $r_{\rm c}=496$ Mm is the radius of the base of the
300: convection zone, $\Delta r$ is the thickness of the return flow, and $A$
301: is the maximum amplitude of the flow (Giles' model B can be roughly
302: approximated with $A \approx 3$~m s$^{-1}$ and $\Delta r \approx 60$ Mm).
303: Figure~\ref{signal} shows travel-time differences, for two
304: focus depths, as a function of $\Delta r$, predicted from the two methods.
305: It is noteworthy that the times computed under the Born approximation
306: can differ substantially from those than predicted by the ray approximation.
307: Much of the sensitivity in the Born approximation lies below the lower
308: turning point of the corresponding ray.
309: In addition, for a sufficiently thick return flow the Born travel-time
310: shifts are greater for the shallower
311: focus depth than for the deeper focus depth.
312: The deeper measurements use waves of higher phase speed, which undergo a
313: smaller phase shift in a horizontally uniform flow.
314:
315:
316: \section{Discussion} \label{discussion}
317:
318: To estimate the amount of data needed for a detection of
319: the return flow, we conservatively assume that
320: a successful detection requires a signal-to-noise ratio
321: (SNR) of three for travel-time measurements near the
322: base of the convection zone. Thus for the types of return flows
323: shown in Figure~\ref{noise_vs_depth}, we require a measurement
324: precision on the order of .01 seconds.
325: Given a random noise of 0.6 seconds for a single day for
326: a HH measurement over a $15^\circ$ strip (Figure~\ref{noise_vs_depth}),
327: it is apparent that at least 12 years of uninterrupted data
328: is needed for a detection of a mean return flow with characteristics
329: similar to Giles' model B using the results of the ray approximation
330: (i.e.\ a travel-time shift of 0.027 seconds), while the
331: more confined flows to the left of the vertical line would require on the order of
332: hundreds of years. These estimates are derived assuming the
333: error in the mean flows decreases with the square root of the
334: length of the time series \citep{Gizon2004}.
335: Combining data from both hemispheres and sacrificing some
336: latitudinal resolution (e.g.\ to $30^\circ$) it should be
337: possible to (roughly) halve the required duration.
338: We therefore tentatively conclude that a 3$\sigma$ detection
339: of a return flow of a magnitude similar or smaller than
340: Giles' model B is possible only with an amount of data
341: comparable or greater than a solar cycle.
342:
343: The additional
344: noise contributions or systematic errors due to analysis
345: artifacts or details of the modeling
346: procedures are not considered here, so that the
347: results represent a ``best-case'' scenario.
348: We anticipate that high-quality measurements spanning a range of
349: depths will actually
350: be needed to construct a model of the flow.
351: Our emphasis on the SNR of the individual measurements
352: is based on the relative ``completeness''
353: of lateral-vantage HH in that a single ingression--egression
354: correlation efficiently samples
355: and combines most of the waves propagating through a particular
356: depth. We therefore assume that the uncertainty in the inferred
357: flow at the base of the convection zone will be dominated by
358: the random noise contribution to measurements focused at that position.
359:
360: It is worthwhile to compare our results with uncertainties in
361: other helioseismic measurements of flows near the base of the
362: convection zone. From inversions of global-mode frequency splittings,
363: \cite{Howe2000} derive 1$\sigma$ errors of inverted rotation
364: rates at a depth of 195 Mm of about 1 nHz (corresponding to 4.4 m s$^{-1}$)
365: for 72-day sets of MDI observations with an averaging kernel
366: approximately 50 Mm
367: wide (FWHM) in depth and about $15^\circ$ wide in latitude.
368: A 3$\sigma$ detection of an average 1.5 m s$^{-1}$ zonal flow over this range
369: in depth and latitude would therefore require about 15 years of
370: frequency-splitting measurements which agrees very well with the estimate presented here.
371:
372: Although direct measurements of temporal variability of the
373: return component over timescales equal to or less than a solar
374: cycle appear unlikely, we note that some inferences about
375: variability could be made using shallower flow measurements
376: and assuming mass conservation, as did \citet{Giles2000}.
377: Regardless of possible temporal variability,
378: it is likely well worth the effort to
379: carry out analyses and modeling of existing decade-length
380: helioseismic observations to either detect or constrain
381: the mean return component. However, we note that there are
382: a considerable number of potential qualifications to the rather
383: simple estimates derived here. Some of these involve issues of the resolution
384: of the models (e.g.\ the ability to infer
385: the existence of multiple meridional cells or sharp gradients
386: of flow in latitude or depth).
387: In addition, we expect considerable challenges to the probing of
388: any flows in the deep convection zone raised by the need to
389: understand and remove possible artifacts and systematic effects.
390: Some of these effects are visible as systematic differences
391: between inferred flows using separate but contemporary datasets
392: \citep{Gonzalez2006}. These differences are often
393: of the order of a few m s$^{-1}$ which, while troublesome
394: for probing even the near-surface layers, are clearly
395: disastrous for the unambiguous identification of a return
396: flow of comparable or smaller magnitude.
397: The critical need for multiple sources of long-duration helioseismic
398: observations combined with careful artifact identification and
399: correction procedures
400: in reducing both systematic effects and random
401: noise should be clear.
402:
403: \acknowledgments
404: We appreciate useful discussions with M. Woodard.
405: DCB and ACB are supported by funding through NASA contract NNH05CC76C and
406: NSF grant AST-0406225, and a subcontract through the
407: NASA sponsored HMI project at Stanford University awarded to NWRA.
408:
409: \begin{thebibliography}{}
410:
411: \bibitem[Birch \& Gizon(2007)]{Birch2007} Birch, A.~C. \& Gizon, L.\
412: 2007, Astron. Nachr., 328, 228
413:
414: \bibitem[Birch et al.(2004)]{Birch2004} Birch, A.~C., Kosovichev, A.~G.,
415: Duvall, T.~L., Jr.\ 2004, \apj, 608, 580
416:
417: \bibitem[Braun et al.(2004)]{Braun2004} Braun, D.~C., Birch,
418: A.~C., \& Lindsey, C.\ 2004, ESA SP-559: SOHO 14 Helio- and
419: Asteroseismology: Towards a Golden Future, 14, 337
420:
421: \bibitem[Braun et al.(2007)]{Braun2007} Braun, D.~C., Birch,
422: A.~C., Benson, D., Stein, R.~F., \& Nordlund, {\AA}.\ 2007, \apj, 669, 1395
423:
424: \bibitem[Braun \& Fan(1998)]{Braun1998} Braun, D.~C., \& Fan, Y.\
425: 1998, \apjl, 508, L105
426:
427: \bibitem[Braun \& Lindsey(2003)]{Braun2003} Braun, D.~C., \&
428: Lindsey, C.\ 2003, ESA SP-517: GONG+ 2002.~Local and Global
429: Helioseismology: the Present and Future, 12, 15
430:
431: \bibitem[Chou \& Ladenkov(2005)]{Chou2005} Chou, D.-Y., \&
432: Ladenkov, O.\ 2005, \apj, 630, 1206
433:
434: \bibitem[Choudhuri et al.(1995)]{Choudhuri1995} Choudhuri, A.~R., Schussler, M.,
435: \& Dikpati, M.\ 1995, \aap, 303, L29
436:
437: \bibitem[Christensen-Dalsgaard et al.(1989)]{JCD89} Christensen-Dalsgaard, J.,
438: Thompson, M.~J., Gough, D.~O.\ 1989, MNRAS, 238, 481
439:
440: \bibitem[Christensen-Dalsgaard et al.(1996)]{Jcd1996}
441: Christensen-Dalsgaard, J., et al.\ 1996, Science, 272, 1286
442:
443: \bibitem[Dikpati \& Gilman(2001)]{Dikpati2001}
444: Dikpati, M., \& Gilman, P.~A.\ 2001, \apj, 559, 428
445:
446: \bibitem[Dikpati \& Gilman(2006)]{Dikpati2006}
447: Dikpati, M., \& Gilman, P.~A.\ 2006, \apj, 649, 498
448:
449: \bibitem[Dikpati \& Gilman(2007)]{Dikpati2007}
450: Dikpati, M., \& Gilman, P.~A.\ 2007, \solphys, 241, 1
451:
452: \bibitem[Giles(2000)]{Giles2000} Giles, P.~M.\ 2000,
453: Ph.D.~Thesis, Stanford University
454:
455: \bibitem[Giles et al.(1998)]{Giles1998} Giles, P.~M., Duvall,
456: T.~L., Jr., \& Scherrer, P.~H.\ 1998, Structure and Dynamics of the
457: Interior of the Sun and Sun-like Stars, 418, 775
458:
459: \bibitem[Giles et al.(1997)]{Giles1997} Giles, P.~M.\, Duvall,
460: T.~L., Jr., Scherrer, P.~H., \& Bogart, R.~S.\ 1997, \nat, 390, 52
461:
462: \bibitem[Gizon \& Birch(2002)]{Gizon2002} Gizon, L. \& Birch, A.C.\
463: 2002, \apj, 571, 966
464:
465: \bibitem[Gizon \& Birch(2004)]{Gizon2004} Gizon, L., \& Birch, A.~C.\ 2004, \apj, 614, 472
466:
467: \bibitem[Glatzmaier \& Gilman(1982)]{Glatzmaier1982} Glatzmaier, G.~A.,
468: \& Gilman, P.~A.\ 1982, \apj, 256, 316
469:
470: \bibitem[Gonz{\'a}lez Hern{\'a}ndez et al.(2008)]{Gonzalez2008}
471: Gonz{\'a}lez Hern{\'a}ndez, I., Kholikov, S., Hill, F., Howe, R.,
472: \& Komm, R.\ 2008, ArXiv e-prints, 808, arXiv:0808.3606
473:
474: \bibitem[Gonz{\'a}lez Hern{\'a}ndez et al.(2006)]{Gonzalez2006}
475: Gonz{\'a}lez Hern{\'a}ndez, I., Komm, R., Hill, F., Howe, R., Corbard, T.,
476: \& Haber, D.~A.\ 2006, \apj, 638, 576
477:
478: \bibitem[Gonz{\'a}lez Hern{\'a}ndez et al.(1999)]{Gonzalez1999}
479: Gonz{\'a}lez Hern{\'a}ndez, I., Patr{\'o}n, J., Bogart, R.~S.,
480: \& The SOI Ring Diagram Team 1999, \apjl, 510, L153
481:
482: \bibitem[Haber et al.(2002)]{Haber2002} Haber, D.~A., Hindman,
483: B.~W., Toomre, J., Bogart, R.~S., Larsen, R.~M., \& Hill, F.\ 2002, \apj,
484: 570, 855
485:
486: \bibitem[Hathaway(1996)]{Hathaway1996} Hathaway, D.~H.\ 1996, \apj,
487: 460, 1027
488:
489: \bibitem[Hathaway et al.(2003)]{Hathaway2003} Hathaway, D.~H.,
490: Nandy, D., Wilson, R.~M., \& Reichmann, E.~J.\ 2003, \apj, 589, 665
491:
492: \bibitem[Hathaway et al.(2004)]{Hathaway2004} Hathaway, D.~H.,
493: Nandy, D., Wilson, R.~M., \& Reichmann, E.~J.\ 2004, \apj, 602, 543
494:
495: \bibitem[Howe et al.(2000)]{Howe2000} Howe, R.,
496: Christensen-Dalsgaard, J., Hill, F., Komm, R.~W., Larsen, R.~M., Schou, J.,
497: Thompson, M.~J., \& Toomre, J.\ 2000, Science, 287, 2456
498:
499: \bibitem[Hughes \& Thompson(2003)]{Hughes2003} Hughes, S.~J., \&
500: Thompson, M.~J.\ 2003, GONG+ 2002.~Local and Global Helioseismology: the
501: Present and Future, 517, 307
502:
503: \bibitem[Lindsey \& Braun(1997)]{Lindsey1997} Lindsey, C. \& Braun, D.~C.\ 1997,
504: \apj, 485, 895
505:
506: \bibitem[Lindsey \& Braun(2000)]{Lindsey2000} Lindsey, C. \& Braun, D.~C.\ 2000,
507: \solphys, 192, 261
508:
509: \bibitem[Lindsey \& Braun(2004)]{Lindsey2004} Lindsey, C. \& Braun, D.~C.\ 2004,
510: \apjs, 155, 209
511:
512: \bibitem[Mitra-Kraev \& Thompson(2007)]{Mitra-Kraev2007} Mitra-Kraev, U., \& Thompson, M.~J.\ 2007, Astronomische Nachrichten, 328, 1009
513:
514: \bibitem[Scherrer et al.(1995)]{Scherrer1995} Scherrer, P.~H. et al.\ 1995,
515: \solphys, 162, 129
516:
517: \bibitem[{\v S}vanda et al.(2007)]{Svanda2007} {\v S}vanda, M.,
518: Zhao, J., \& Kosovichev, A.~G.\ 2007, \solphys, 241, 27
519:
520: \bibitem[Wang et al.(2002)]{Wang2002} Wang, Y.-M., Lean, J.,
521: \& Sheeley, N.~R., Jr.\ 2002, \apjl, 577, L53
522:
523: \bibitem[Wang et al.(1991)]{Wang1991} Wang, Y.-M., Sheeley,
524: N.~R., Jr., \& Nash, A.~G.\ 1991, \apj, 383, 431
525:
526: \bibitem[Zhao \& Kosovichev(2004)]{Zhao2004} Zhao, J., \&
527: Kosovichev, A.~G.\ 2004, \apj, 603, 776
528:
529: \end{thebibliography}
530:
531: \begin{deluxetable}{ccccc}
532: \tablecolumns{3}
533: \tablewidth{0pc}
534: \tablecaption{Pupil size and mode degrees\label{tbl-1}}
535: \tablehead{
536: \colhead{Depth} & \colhead{Pupil radii} & \colhead{$\ell$ @ 4mHz}\\
537: \colhead{(Mm)} & \colhead{(degrees)} & \colhead{ }\\
538: }
539: \startdata
540: 30 & 1.3 - 10.5 & 166 - 235 \\
541: 37 & 1.7 - 12.9 & 145 - 205 \\
542: 45 & 2.0 - 15.4 & 128 - 180 \\
543: 54 & 2.4 - 18.2 & 113 - 159 \\
544: 64 & 2.9 - 21.1 & 100 - 142 \\
545: 76 & 3.4 - 24.1 & 89 - 127 \\
546: 88 & 4.0 - 26.5 & 81 - 114 \\
547: 100 & 4.5 - 29.2 & 73 - 104 \\
548: 114 & 5.2 - 32.0 & 66 - 93 \\
549: 130 & 5.9 - 35.2 & 59 - 84 \\
550: 150 & 6.8 - 41.4 & 52 - 74 \\
551: 170 & 7.7 - 48.1 & 46 - 66 \\
552: 190 & 8.6 - 54.9 & 41 - 58 \\
553: 200 & 9.1 - 57.1 & 39 - 55 \\
554: \enddata
555: \end{deluxetable}
556:
557: \begin{figure}[htbp]
558: \epsscale{1.0}
559: \plotone{f1.eps}
560: \caption{Maps of the mean (panels a-c) and standard deviation
561: (panels d-f) of north-south travel-time differences for twenty
562: consecutive 24-hr
563: measurements. From left to right the focus depths of
564: the measurements are 30, 100, and 200 Mm respectively.
565: }
566: \label{maps}
567: \end{figure}
568:
569: \begin{figure}[htbp]
570: \epsscale{1.}
571: \plotone{f2.eps}
572: \caption{
573: The standard deviation (over twenty consecutive 24-hr measurements)
574: of the mean north-south travel-time difference
575: averaged over a strip
576: of the Sun spanning 15 degrees in latitude and 120 degrees in central
577: meridian distance.
578: Different colors indicate different latitudes of the center of the strip,
579: while solid (dashed) lines indicate the southern (northern) hemisphere.
580: In general, the mean-standard-deviation increases with the depth of
581: the focus, and also increases at high latitudes.
582: }
583: \label{noise_vs_depth}
584: \end{figure}
585:
586: \begin{figure}[htbp]
587: \epsscale{1.}
588: \plotone{f3.eps}
589: \caption{
590: Expected north--south travel-time difference
591: as functions of the width of a
592: hypothesized meridional return flow at the base of
593: the convection zone with a peak value of 3 m s$^{-1}$(see text).
594: The flow is set to zero in the radiative zone.
595: The red (black) lines show the results of a Born (ray)
596: approximation calculation, and the
597: the solid (dashed) lines show the results for focus depths of
598: 200 (170) Mm below the surface.
599: The vertical line indicates the width which roughly corresponds to
600: Giles' model B.
601: }
602: \label{signal}
603: \end{figure}
604:
605: \end{document}
606: