1: %\documentclass[12pt,preprint]{aastex}

2: \documentclass{emulateapj}  

3: 

4: \usepackage{graphicx} 

5: \slugcomment{Accepted in ApJ Letters}

6: \shorttitle{Quiet sun magnetic fields from space-borne observations} 

7: \shortauthors{Orozco Su\'arez et al.\/}

8: 

9: \begin{document}

10: 

11: 

12: \title{Quiet Sun magnetic fields from space-borne observations: simulating 

13: Hinode's case} 

14: \author{D.\ Orozco Su\'arez, L.R.\ Bellot Rubio, and J.C.\ del Toro Iniesta}

15: 

16: \affil{Instituto de Astrof\'{\i}sica de Andaluc\'{\i}a (CSIC), Apdo.\ 3004, 

17: 18080 Granada, Spain; orozco@iaa.es, lbellot@iaa.es, jti@iaa.es}

18: 

19: \begin{abstract} 

20: We examine whether or not it is possible to derive the field strength

21: distribution of quiet Sun internetwork regions from very high spatial

22: resolution polarimetric observations in the visible.  In particular,

23: we consider the case of the spectropolarimeter attached to the Solar

24: Optical Telescope aboard Hinode.  Radiative magneto-convection

25: simulations are used to synthesize the four Stokes profiles of the

26: \ion{Fe}{1} 630.2~nm lines. Once the profiles are degraded to a

27: spatial resolution of 0\farcs32 and added noise, we infer the

28: atmospheric parameters by means of Milne-Eddington inversions. The

29: comparison of the derived values with the real ones indicates that the

30: visible lines yield correct internetwork field strengths and magnetic

31: fluxes, with uncertainties smaller than $\sim$150~G, when a stray

32: light contamination factor is included in the inversion.  Contrary to

33: the results of ground-based observations at 1\arcsec\/, weak fields

34: are retrieved wherever the field is weak in the simulation.

35: \end{abstract}

36: 

37: \keywords{Sun: magnetic fields -- Sun: photosphere 

38: -- Instrumentation: high angular resolution}

39: 

40:   \section{Introduction}

41:   \label{sec:intro}

42: The characterization of quiet sun internetwork (IN) fields is an important

43: issue in solar physics. Polarimetric measurements of the visible \ion{Fe}{1}

44: lines at 630.2~nm and the near-infrared \ion{Fe}{1} lines at 1565~nm have been

45: used to advance our knowledge of IN magnetism, but no consensus has been

46: reached yet.  While the analysis of the visible lines suggest a predominance

47: of kG field strengths and small filling factors (S\'anchez Almeida \& Lites

48: 2000; Dom\'{\i}nguez Cerde\~na et al.\ 2003; Socas-Navarro \& Lites 2004), the

49: near-infrared lines indicate that most fields have hG strengths with

50: larger filling factors (Lin 1995; Lin \& Rimele 1999; Khomenko et al.\ 2003;

51: Mart\'{\i}nez Gonz\'alez et al.\ 2006; Dom\'{\i}nguez Cerde\~na et al.\ 2006).

52: 

53: Attempts to reconcile these contradictory results have argued that visible and

54: IR lines sample different magnetic structures in the resolution element

55: (S\'anchez Almeida \& Lites 2000; Socas-Navarro \& S\'anchez Almeida 2003) or

56: that noise affects the visible lines more dramatically than the IR lines

57: (Be\-llot Rubio \& Collados 2003). On the other hand, Mart\'{\i}nez Gonz\'alez

58: et al.\ (2006) have convincingly demonstrated that it is not possible to

59: obtain reliable IN field strengths from the \ion{Fe}{1} 630.2~nm lines due to

60: crosstalk with thermodynamical parameters at 1\arcsec\/ resolution.

61: %In

62: %view of these difficulties, some authors have explored new ways to

63: %derive the fields of IN regions (Asensio Ramos et al.\ 2006; L\'opez

64: %Ariste et al.\ 2006: Trujillo Bueno et al.\ 2004).

65: 

66: Different techniques have been proposed to study IN fields, but the diagnostic

67: potential of high spatial resolution observations in the absence of

68: atmospheric seeing has only been explored by Khomenko et al.\ (2004).

69: %spatial resolution measurements in the absence of atmospheric seeing. 

70: The spectropolarimeter (SP; Lites et al.\ 2001) of the Solar Optical Telescope

71: (SOT) aboard Hinode (Ichimoto et al.\ 2005) is already providing nearly

72: diffraction-limited observations of the solar photosphere, and upcoming

73: instruments will do so in the future (e.g., IMaX aboard SUNRISE or VIM 

74: aboard Solar Orbiter). Thus, there is a clear need to assess whether 

75: reliable IN field strengths can be derived from space-borne observations.

76: 

77: Here we address this question by simulating and analyzing

78: Hinode/SP measurements. Radiative magneto-convection simulations are

79: used to synthesize the Stokes profiles of the \ion{Fe}{1}

80: 630.2~nm lines. The profiles are degraded to the nearly

81: diffraction-limited resolution of 0\farcs32 achieved by

82: Hinode/SP. After adding noise to the Stokes spectra, the atmospheric

83: parameters are inferred by means of inversion techniques. By comparing

84: the derived values with the real ones we determine the errors in field

85: strength and magnetic flux to be expected from the analysis of Hinode

86: data. Our main result is that Milne-Eddington (ME) inversions of the

87: visible 630.2~nm lines yield correct IN field strengths

88: with uncertainties smaller than 150~G for the whole range of strengths

89: from $\sim$0.1 to 1~kG.  If internetwork fields are hG fields, then

90: simple ME inversions of Hinode/SP measurements will result in sub-kG

91: field strength distributions, contrary to what is obtained from

92: current ground-based observations.

93: 

94: %

95: %________________________________________________________________

96: 

97: 

98: \section{MHD simulations, spectral synthesis, and image degradation}

99: \label{sec:simul}

100: 

101: To describe the sun in the more realistic way possible we use

102: radiative MHD simulations by V\"ogler et al.\ (2005).  Specifically,

103: we take three snapshots from different simulation runs representing

104: very quiet, unipolar internetwork and network regions with average

105: unsigned fluxes of 10, 50, and 200~Mx~cm$^{-2}$, respectively. The

106: horizontal and vertical extents of the computational box are 6 and

107: 1.4~Mm. The synthesis of the Stokes spectra of the two \ion{Fe}{1}

108: lines is carried out using the SIR code (Ruiz Cobo \& del Toro Iniesta

109: 1992). The spectral region is sampled at 113 wavelength positions in

110: steps of 2.15~pm, following the Hinode/SP normal map mode (for

111: details, see Shimizu 2004). The atomic parameters have been taken from

112: the VALD database (Piskunov et al.\ 1995).

113: 

114: %FIG 1 POS

115: \begin{figure*}

116: \centering 

117: \epsscale{0.38} 

118: \plotone{f1.ps} 

119: \plotone{f2.ps}

120: \plotone{f3.ps}

121: \caption{{\em Left and middle:} Continuum intensity maps for the

122: simulation snapshot with average unsigned flux of 10~Mx~cm$^{-2}$ 

123: and for the data spatially degraded considering telescope diffraction 

124: and pixel size. Color scales are the same in the two maps. The contrast 

125: varies from 13.7\% in the original image to 8.5\% in the degraded one. 

126: {\em Right:} MTF of the detector (dotted line), diffraction limited MTF 

127: (dashed line), and combination of both effects (solid line).}

128: \label{fig:fig1}

129: \end{figure*}

130: 

131: The aperture of Hinode/SOT is 0.5~m, which operating at 630~nm provides a

132: spatial resolution of $\sim$0\farcs26 (equivalent to $\sim$\,190~km on the

133: solar surface). The sampling interval in the MHD simulations is 0\farcs0287,

134: implying a spatial resolution of 0\farcs057 (41.6~km). Thus, in order to

135: simulate SOT observations, the synthetic Stokes profiles derived from the MHD

136: model have been spatially degraded by telescope diffraction. In addition, we

137: have considered the extra loss of contrast caused by the integration of the

138: signal in the detector. Finally, the images are rebinned to the SP CCD pixel

139: size of 0$\farcs$16$\times$ 0$\farcs$16.  Figure~\ref{fig:fig1} shows

140: continuum intensity maps for the original and the spatially degraded data. 

141: The degradation process reduces the rms contrast from $\sim$14\% to $\sim$9\% 

142: in the continuum. The pixelation of the CCD is noticeable in the degraded

143: image. The right panel of Fig.~\ref{fig:fig1} shows Modulation Transfer

144: Functions (MTFs) describing the filtering of spectral components induced 

145: by telescope diffraction and pixelation effects in the CCD.  Note the

146: modification of the effective MTF caused by the central obscuration of the

147: entrance pupil.

148: 

149: %FIG 2 POS

150: %figure

151: \begin{figure}

152: \centering

153: \epsscale{0.8}

154: \plotone{f4.ps}

155: \caption{Comparison between the average Stokes $I$ profile from the

156: spatially degraded data (solid) and the FTS spectral atlas

157: (dashed). Both continua are normalized to unity.}

158: \label{fig:fig2}

159: \end{figure}

160: 

161: Figure~\ref{fig:fig2} compares the average Stokes $I$ profiles from the

162: spatially degraded data and the NSO Fourier Transform Spectrometer Atlas of

163: the quiet Sun. Both spectra are very similar, with only small differences in

164: the line core and wings of \ion{Fe}{1} 630.1~nm. Note that the lack of a

165: temporal average exclude, for instance, the effect of the 5-min oscillation 

166: in the simulated profile. To reproduce an Hinode/SP observation we have also

167: convolved the profiles with a Gaussian of 30 m{\AA}~FWHM to account for the

168: spectral resolving power of the spectrograph and have added noise at

169: the level of 10$^{-3}$ of the continuum intensity $I_{\rm c}$ 

170: (the polarimetric sensitivity of standard Hinode/SP measurements).

171: 

172: %

173: %________________________________________________________________

174: 

175: \section{Inversion}

176: 

177: To derive the magnetic field strength from the simulated profiles we

178: use a least-square inversion technique based on ME atmospheres.  ME

179: inversions represent the best option to interpret the measurements if

180: one is not interested in vertical gradients of the physical

181: quantities.  They are simple and often provide reasonable averages of

182: the atmospheric parameters over the line formation region (Westendorp

183: Plaza et al.\ 2001; Bellot Rubio 2006).

184: 

185: We apply the ME inversion to the \ion{Fe}{1} 630.15~nm and \ion{Fe}{1} 

186: 630.25~nm lines simultaneously. A total of 9 free parameters are 

187: determined (S$_0$, S$_1$, $\eta_0$, $\Delta\lambda_D$, $a$, $B$, 

188: $\gamma$, $\chi$, and v$_\mathrm{LOS}$; for the meaning of the symbols 

189: see, e.g., Orozco Su\'arez \& del Toro Iniesta 2007). 

190: No additional broadening of the profiles by macroturbulence 

191: or microturbulence is allowed. 

192: %given the very high spatial and temporal resolution of the

193: %observations.

194: %We have carried out two different inversions to

195: %determine the atmospheric parameters. Both of them use a simple

196: %one-component model. We first invert the profiles with noise added at

197: %the level of 10$^{-3}$ considering zero stray light contamination. In

198: %the second inversion we fit the noisy profiles considering non-zero

199: %stray light contaminations. 

200: Three different inversions are performed to derive the atmospheric

201: parameters. All of them use a simple one-component model, i.e., a

202: laterally homogeneous magnetic atmosphere occupying the whole resolution

203: element. We first invert the profiles in the absence of noise, and

204: then with noise added at the level of 10$^{-3} \, I_{\rm c}$. In the

205: last inversion, the noisy profiles are fitted considering non-zero

206: stray light contaminations factors.

207: %%%%%%%%%%%%%%%%%%

208: The stray light profile is evaluated individually for each pixel by 

209: averaging the Stokes $I$ profiles within a box 1\arcsec\/-wide centered 

210: on the pixel. For all inversions we use the same initial guess model, 

211: allowing a maximum of 300 iterations. The initial field strength is 100~G. 

212: %We analyze only pixels whose Stokes $Q$, $U$ or $V$ amplitudes 

213: %exceed three times the noise level. 

214: %Once the profiles are 

215: %inverted, we compare the inferred maps with the ``real'' model

216: %atmospheres.

217: 

218: %

219: %________________________________________________________________

220: 

221: \section{Results}

222: \label{sec:res}

223: 

224: Figure~\ref{fig:mapas} shows the vector magnetic field (strength,

225: inclination, and azimuth) retrieved from the inversions of the Stokes

226: profiles. The first column displays a cut of the simulation snapshot

227: with average flux density of 10~Mx~cm$^{-2}$ at optical depth ${\rm

228: log} \, \tau=-2$.  The second and third columns contain the results of

229: the ME inversions of the spatially degraded profiles in the absence of

230: noise and the specific case of a SNR of 1000, respectively. Finally,

231: the fourth column shows the atmospheric parameters derived from the

232: noisy profiles accounting for stray light contamination. White regions

233: represent pixels which have not been inverted because of their small

234: polarization signals (we only consider pixels whose Stokes $Q$, $U$ or

235: $V$ amplitudes exceed three times the noise level).

236: 

237: %FIG 3 POS

238: %figure

239: \begin{figure*}

240: \centering

241: \epsscale{1.1}

242: \plotone{f5.ps}

243: \vspace{-2em}

244: \caption{{\em Left}: Cuts at optical depth ${\rm log} \, \tau =-2$ of

245: the model atmospheres provided by the MHD simulation with averaged

246: unsigned flux of 10~Mx~cm$^{-2}$. {\em Second column}: Maps of the

247: physical quantities retrieved from the ME inversion of the profiles

248: with no noise. {\em Third column}: Maps retrieved from the ME

249: inversion of the profiles with SNR of 1000 and no stray light

250: contamination. {\em Fourth column:} Same as before, but accounting for

251: stray light. From top to bottom: magnetic field strength, inclination,

252: and azimuth.}

253: \label{fig:mapas}

254: \end{figure*}

255: 

256: Over the granules, the magnetic field strength is very weak and the

257: polarization signals are buried in the noise. These pixels represent

258: $\sim$55\% of the total area (white regions in Fig.~\ref{fig:mapas}).

259: The stronger fields concentrate in intergranular regions. In those

260: regions, the magnetic structures inferred from the inversion have

261: bigger sizes than the real ones, i.e., they appear ''blurred''. This 

262: is caused by the degradation of the images due to telescope 

263: diffraction and CCD pixel size.  The field inclination and 

264: azimuth structures resulting from the inversion are blurred as 

265: well. The azimuth values are rather uncertain because of the tiny 

266: linear polarization signals produced by the weak fields of the 

267: simulations.

268: 

269: Figure \ref{fig:mapaszoom} is a close up of small features observed in

270: intergranular regions. Note that each Hinode/SP pixel of $0\farcs16

271: \times 0\farcs16$ corresponds to 36 pixels in the simulation, hence

272: they usually contain a broad distribution of magnetic field

273: strengths. When we consider that the polarization signal is produced

274: by a single magnetic component within the resolution element and no

275: stray light is allowed for, the inferred field strengths are smaller

276: than those in the model, so the field is underestimated (middle panels

277: of Fig.~\ref{fig:mapaszoom}). If one accounts for stray light

278: contamination the inferred fields become stronger (right panels), but

279: also noisier due to the increased number of free parameters.

280: 

281: %FIG 4 POS

282: %figure

283: \begin{figure}[!t]

284: \centering

285: \epsscale{1.1}

286: \plotone{f6.ps}

287: \plotone{f7.ps}

288: \caption{{\em Left:} Field strengths at ${\rm log} \, \tau =-2$ in the

289: MHD simulations with 10~Mx~cm$^{-2}$ (top) and 50~Mx~cm$^{-2}$

290: (bottom). {\em Middle}: Field strengths derived from the ME inversion

291: of the spatially degraded Stokes profiles with SNR 1000 and no stray

292: light contamination. {\em Right}: Field strengths from the ME

293: inversion accounting for stray light contamination. }

294: \label{fig:mapaszoom}

295: \end{figure}

296: 

297: To analyze the results in a more quantitative way we calculate the mean 

298: and rms values of the errors. We define the error as the difference

299: between the inferred and the real parameters at optical depth ${\rm

300: log} \, \tau=-2$. Since one pixel of the degraded data corresponds 

301: to 36 pixels in the simulations, we compare each inverted pixel with 

302: the mean of the corresponding 36 pixels in the original map. 

303: Figure~\ref{fig:fig5} shows the mean and rms errors of the field 

304: strength resulting from the inversion without accounting for stray

305: light (top left panel). It is clear that fields above $\sim$100~G are

306: underestimated, with rms errors smaller than $\sim$150~G in the whole

307: range of strengths. The results are similar for the magnetic flux

308: density (top right panel). The inversion considering stray light

309: contamination yields much better inferences, as can be seen in the

310: bottom panels of Fig.~\ref{fig:fig5}. The field strength and flux 

311: are slightly overestimated for weak fields, but the rms errors 

312: do not exceed 150~G in any case.

313: 

314: %FIG 5 POS

315: %figure

316: \begin{figure}

317: \centering 

318: \epsscale{1.1}

319: \plotone{f8.ps}   

320: \vspace*{.5em}

321: \caption{{\em Top:} Mean (solid) and rms (dashed) errors of the field

322: strength (left) and flux (right) derived from the inversion of the

323: profiles with SNR=1000 assuming a single magnetic atmosphere and no

324: stray light contamination. {\em Bottom:} Same as before but accounting

325: for stray light contamination.}

326: \label{fig:fig5}

327: \end{figure}

328: 

329: %

330: %________________________________________________________________

331: 

332: \section{Discussion}

333: 

334: Previous analyses of visible (630.2~nm) and near-infrared (1565~nm)

335: neutral iron lines do not agree on the distribution of field strengths

336: in IN regions. In particular, the visible lines systematically deliver

337: kG field strengths and small filling factors, while the near-infrared

338: lines suggest a predominance of hG fields.

339: 

340: Here we have shown that ME inversions of the \ion{Fe}{1}~630~nm lines at

341: spatial resolutions of 0\farcs32 (the case of Hinode/SP) underestimate the

342: magnetic field strength by some hundred G if no stray light contamination is

343: included.  When stray light is accounted for, ME inversions are able

344: to recover any magnetic field above 100~G with remarkable accuracy.

345: 

346: Interestingly, we always derive weak fields from the simulated

347: Hinode/SP observations where the field in the MHD model is

348: weak. Likewise, pixels assigned strong fields by the ME inversion

349: actually correspond to strong fields in the MHD model. This is in

350: sharp contrast with the results of Mart\'{\i}nez Gonz\'alez et al.\

351: (2006), who always infer kG fields from the \ion{Fe}{1} 630.2~nm 

352: lines observed in the IN at resolutions of 1\arcsec\--1\farcs5 when 

353: the inversion is initialized with strong fields. The difference is

354: probably due to: (a) our significantly higher spatial resolution,

355: which narrows the range of field strengths present in the pixel; (b)

356: the fact that we do not employ two-component atmospheres, micro- or

357: macro-turbulent velocities, which reduces the degrees of freedom of

358: the solution; and (c) the simple description of the thermodynamics

359: provided by the ME model which, contrary to the atmosphere used by

360: Mart\'{\i}nez Gonz\'alez et al.\ (2006), does not allow to compensate

361: for incorrect magnetic parameters.  On the other hand, our results

362: seem to contradict the conclusions of Bellot Rubio \& Collados

363: (2003). However, the signals considered here are larger by a factor 

364: of $\sim$10 due to the much higher angular resolution (which implies

365: larger filling factors). Under these conditions, noise does not

366: significantly affect the field strengths derived from the visible

367: lines (see Fig.\ 5 in Bellot Rubio \& Collados 2003). 

368: 

369: We caution that the results of this Letter may only be valid as 

370: long as the MHD simulations provide a realistic description of the Sun.  

371: The performance of ME inversions could be different if the magnetic field 

372: is structured on scales much smaller than $\sim$0\farcs3. For the moment,

373: however, there is no compelling evidence that tiny magnetic elements 

374: exist in the quiet solar photosphere.

375: 

376: 

377: %________________________________________________________________

378: 

379:   \section{Conclusions}

380:   \label{sec:con}

381: 

382: 

383: Our analysis suggests that Hinode/SP observations will make it possible to

384: determine the real distribution of field strengths in quiet Sun internetwork

385: regions. Simple one-component Milne-Eddington inversions without macro and

386: microturbulence seem appropriate to achieve that goal. However, it will be

387: essential to account for the degradation of the image induced by telescope

388: diffraction and detector pixel size. The work presented here shows that the

389: effects of the degradation can be modeled sufficiently well including a 

390: stray light contamination factor in the inversion.

391: 

392: 

393: 

394: %

395: %________________________________________________________________

396: 

397: \acknowledgments We thank A.\ V\"ogler and M.\ Sch\"ussler for making

398: their MHD simulations available and answering our questions about

399: them. The program used to degrade the synthetic Stokes profiles was

400: written by J.A.\ Bonet and S.\ Vargas. This work has been partially

401: funded by the Spanish Mi\-nisterio de Educaci\'on y Ciencia through

402: project ESP2003-07735-C04-03 (in\-clu\-ding European FEDER funds) and

403: {\em Programa Ram\'on y Cajal}.

404: 

405: %

406: %________________________________________________________________

407: 

408: \begin{thebibliography}{}

409: 

410: %%\bibitem[Asensio Ramos et al.(2006)]{2006astro.ph.12389A} Asensio Ramos, 

411: %%A., Mart\'{\i}nez Gonz\'alez, M.J., Lopez Ariste, A., Trujillo Bueno, J., \& 

412: %%Collados, M.\ 2007, \apj, submitted (astro-ph/0612389)  %***

413: 

414: 

415: \bibitem[Bellot Rubio(2006)]{spw4} Bellot Rubio, L.~R.\

416: 2006, ASP Conf.\ Ser.: Solar Polarization 4, 358, 107%***

417: 

418: \bibitem[Bellot Rubio \& Collados(2003)]{2003A&A...406..357B} Bellot Rubio, 

419: L.~R., \& Collados, M.\ 2003, \aap, 406, 357 %***

420: 

421: \bibitem[]{} 

422: Dom{\'{\i}}nguez Cerde{\~n}a, I., Kneer, F., \& S{\'a}nchez Almeida, J.\ 

423: 2003, \apjl, 582, L55  %***

424: 

425: \bibitem[Dom{\'{\i}}nguez Cerde{\~n}a et al.(2006)]{2006ApJ...646.1421D} 

426: Dom{\'{\i}}nguez Cerde{\~n}a, I., Almeida, J.~S., \& Kneer, F.\ 2006, \apj, 

427: 646, 1421 %***

428: 

429: %\bibitem[Ichimoto et al.(2004)]{2004SPIE.5487.1142I} Ichimoto, K., et al.\ 

430: %2004, \procspie, 5487, 1142 

431: 

432: \bibitem[Ichimoto \& Solar-B Team(2005)]{2005JKAS...38..307I} Ichimoto, K., 

433: \& Solar-B Team 2005, Journal of Korean Astronomical Society, 38, 307 %***

434: 

435: \bibitem[Khomenko et al.(2003)]{2003A&A...408.1115K} Khomenko, E.~V., 

436: Collados, M., Solanki, S.~K., Lagg, A., \& Trujillo Bueno, J.\ 2003, \aap, 

437: 408, 1115 %***

438: 

439: \bibitem{} Khomenko, E., Collados, M., \& Solanki, S.K.\ 2004, Mem.\ S.A. It., 

440: 75, 282

441: 

442: \bibitem[Lin(1995)]{1995ApJ...446..421L} Lin, H.\ 1995, \apj, 446, 421 %***

443: 

444: \bibitem[Lin \& Rimmele(1999)]{1999ApJ...514..448L} Lin, H., \& Rimmele, 

445: T.\ 1999, \apj, 514, 448 %***

446: 

447: \bibitem[Lites et al.(2001)]{2001ASPC..236...33L} Lites, B.~W., Elmore, 

448: D.~F., \& Streander, K.~V.\ 2001, ASP Conf.~Ser.: Advanced Solar 

449: Polarimetry -- Theory, Observation, and Instrumentation, 236, 33 %***

450: 

451: %%\bibitem[L{\'o}pez Ariste et al.(2006)]{2006A&A...454..663L} L{\'o}pez 

452: %%Ariste, A., Tomczyk, S., \& Casini, R.\ 2006, \aap, 454, 663  %***

453: 

454: \bibitem[Mart{\'{\i}}nez Gonz{\'a}lez et al.(2006)]{2006A&A...456.1159M} 

455: Mart{\'{\i}}nez Gonz{\'a}lez, M.J., Collados, M., \& Ruiz Cobo, B.\ 2006, 

456: \aap, 456, 1159 %***

457: 

458: \bibitem[Orozco Su{\'a}rez \& del Toro Iniesta(2007)]{2007A&A...462.1137O} 

459: Orozco Su{\'a}rez, D., \& del Toro Iniesta, J.C.\ 2007, \aap, 462, 1137 %*****

460: 

461: \bibitem[Piskunov et al.(1995)]{piskunov} Piskunov, N.~E.,

462: Kupka, F., Ryabchikova, T.~A., Weiss, W.~W., \& Jeffery, C.~S.\ 1995,

463: A\&A Supp.\ Ser., 112, 525%***

464: 

465: \bibitem[Ruiz Cobo \& del Toro Iniesta(1992)]{1992ApJ...398..375R} Ruiz 

466: Cobo, B., \& del Toro Iniesta, J.~C.\ 1992, \apj, 398, 375 %***

467: 

468: %\bibitem[Sch{\"u}ssler et al.(2003)]{shussler} Sch{\"u}ssler,

469: %M., Shelyag, S., Berdyugina, S., V{\"o}gler, A., \& Solanki, S.K.\ 2003,

470: %ApJ, 597, L173%***

471: 

472: \bibitem[]{} S\'anchez Almeida, J., \& Lites, B.W.\ 2000, \apj, 532, 1215 %***

473: 

474: \bibitem[Shimizu(2004)]{2004ASPC..325....3S} Shimizu, T.\ 2004, ASP 

475: Conf.~Ser.: The Solar-B Mission and the Forefront of Solar Physics, 

476: 325, 3 %***

477: 

478: \bibitem[]{}

479: Socas-Navarro, H., \& S\'anchez Almeida, J.\ 2003, \apj, 593, 581 %***

480: 

481: \bibitem[]{} Socas-Navarro, 

482: H., \& Lites, B.~W.\ 2004, \apj, 616, 587 %***

483: 

484: %%\bibitem[Trujillo Bueno et al.(2004)]{2004Natur.430..326T} Trujillo Bueno, 

485: %%J., Shchukina, N., \& Asensio Ramos, A.\ 2004, \nat, 430, 326 %***

486: 

487: \bibitem[V{\"o}gler et al.(2005)]{voegler} V{\"o}gler, A.,

488: Shelyag, S., Sch{\"u}ssler, M., Cattaneo, F., Emonet, T., \& Linde,

489: T.\ 2005, A\&A, 429, 335%***

490: 

491: \bibitem[Westendorp Plaza et al.(2001)]{carlos} Westendorp

492: Plaza, C., del Toro Iniesta, J.C., Ruiz Cobo, B., Mart\'{\i}nez Pillet, V., 

493: Lites, B.W., \& Skumanich, A.\ 2001, ApJ, 547, 1130%***

494: 

495: 

496: \end{thebibliography}

497: 

498: 

499: %

500: %________________________________________________________________

501: 

502: \end{document}

503: 

504: 

505: