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

2: %\documentclass[12pt,preprint]{emulateapj}

3: 

4: %% manuscript produces a one-column, double-spaced document:

5: 

6: % \documentclass[manuscript]{aastex}

7: 

8: %% preprint2 produces a double-column, single-spaced document:

9: 

10: \documentclass[preprint2]{aastex}

11: 

12: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

13: % For figures:

14: % Allocate space or include encapsulated PostScript

15: %

16: %\newcommand{\epsfig}[1]{\epsfxsize=8cm \epsfbox{#1}}

17: \newcommand{\epsfig}[1]{\resizebox{\hsize}{!}{\includegraphics{#1}}}

18: % \newcommand{\epsfig}[1]{\picplace{6.5cm}}

19: % A figure caption macro

20: %

21: \newcommand{\figcap}[1]{\caption[Another figure]{#1}}

22: %

23: % Here comes the bibliography macro definitions

24: %

25: %\input{bibmacros.tex}

26: %

27: % Here comes some astronomical special signs

28: % like magnitudes, seconds, minutes and hours and arcsec's arcmin's

29: % and Mv,Mbol,Mk,Teff

30: %

31: \newcommand{\Teff}{\mbox{$T_{\mbox{\scriptsize eff}}\,$}}

32: \newcommand{\Tefffour}{\mbox{$T^{4}_{\mbox{\scriptsize eff}}\,$}}

33: \newcommand{\Teq}{\mbox{$T_{\mbox{\scriptsize eq}}\,$}}

34: \newcommand{\Sv}{\mbox{$S_{V}\,$}}

35: \newcommand{\BCv}{\mbox{$\mbox{B.C.}_{V}\,$}}

36: \newcommand{\geff}{\mbox{$g_{\mbox{\scriptsize eff}}\,$}}

37: \newcommand{\Vrot}{\mbox{$v_{\mbox{\scriptsize rot}}\,$}}

38: \newcommand{\mbol}{\mbox{$m_{\mbox{\scriptsize bol}}\,$}}

39: \newcommand{\mv}{\mbox{$m_{V}\,$}}

40: \newcommand{\mk}{\mbox{$m_{K}\,$}}

41: \newcommand{\Mbol}{\mbox{$M_{\mbox{\scriptsize bol}}\,$}}

42: \newcommand{\Mv}{\mbox{$M_{V}\,$}}

43: \newcommand{\Mk}{\mbox{$M_{K}\,$}}

44: \newcommand{\Mrr}{\mbox{$M_{\mbox{\scriptsize RR}}\,$}}

45: \newcommand{\delMv}{\mbox{$\Delta 

46: 	 M_{\mbox{\scriptsize V,TO}}^{\mbox{\scriptsize RR}}\,$}}

47: \newcommand{\meanMbol}{\mbox{$<\!\!M_{\mbox{\scriptsize bol}}\!\!>\,$}}

48: \newcommand{\meanMv}{\mbox{$<\!\!M_{V}\!\!>\,$}}

49: \newcommand{\meanMk}{\mbox{$<\!\!M_{K}\!\!>\,$}}

50: \newcommand{\Koff}{\mbox{$K_{\mbox{\scriptsize off}}\,$}}

51: \newcommand{\MvRR}{\mbox{$<M_{V}(RR)>\,$}}

52: \newcommand{\MkRR}{\mbox{$<M_{K}(RR)>\,$}}

53: \newcommand{\MbolRR}{\mbox{$<M_{\mbox{\scriptsize bol}}(RR)>\,$}}

54: \newcommand{\MvTO}{\mbox{$M_{V}(TO)\,$}}

55: \newcommand{\RRab}{\mbox{$RR{\mbox{\scriptsize ab}}\,$}}

56: \newcommand{\RRc}{\mbox{$RR{\mbox{\scriptsize c}}\,$}}

57: \newcommand{\RRd}{\mbox{$RR{\mbox{\scriptsize d}}\,$}}

58: \newcommand{\pmag}{\mbox{$\stackrel{\mbox{\scriptsize m}}{\textstyle .}$}}

59: \newcommand{\mags}{\mbox{$^{\mbox{\scriptsize m}}$}}

60: \newcommand{\pday}{\mbox{$\stackrel{\mbox{d}}{\textstyle .}$}}

61: \newcommand{\pperiod}{\mbox{$\stackrel{p}{\textstyle .}$}}

62: \newcommand{\phour}{\mbox{$\stackrel{h}{\textstyle .}$}}

63: \newcommand{\hours}{\mbox{$^{h}$}}

64: \newcommand{\pmin}{\mbox{$\stackrel{m}{\textstyle .}$}}

65: \newcommand{\mins}{\mbox{$^{m}$}}

66: \newcommand{\psec}{\mbox{$\stackrel{s}{\textstyle .}$}}

67: \newcommand{\secs}{\mbox{$^{s}$}}

68: \newcommand{\parcsec}{\mbox{$\stackrel{\prime\prime}{\textstyle .}$}}

69: \newcommand{\arcsecs}{\mbox{$^{\prime\prime}$}}

70: \newcommand{\parcmin}{\mbox{$\stackrel{\prime}{\textstyle .}$}}

71: \newcommand{\arcmins}{\mbox{$^{\prime}$}}

72: \newcommand{\solar}{\mbox{$\odot$}}

73: \newcommand{\Rsolar}{\mbox{$R_{\odot}\,$}}

74: \newcommand{\Msolar}{\mbox{$\vec{M}_{\odot}\,$}}

75: \newcommand{\Lsolar}{\mbox{$L_{\odot}\,$}}

76: \newcommand{\Tsolar}{\mbox{$T_{\odot}\,$}}

77: \newcommand{\FeH}{\mbox{[Fe/H]}\,}

78: \newcommand{\OFe}{\mbox{[O/Fe]}\,}

79: \newcommand{\kms}{\mbox{$\mbox{km\,s}^{-1}$}\,}

80: \newcommand{\Mpc}{\mbox{Mpc}\,}

81: \newcommand{\prMpc}{\mbox{$\mbox{Mpc}^{-1}$}\,}

82: \newcommand{\logg}{\mbox{$\log g$}\,}

83: \newcommand{\loggzero}{\mbox{$\log g_0$}\,}

84: \newcommand{\logP}{\mbox{$\log P$}\ }

85: \newcommand{\mM}{\mbox{$(m-M)$}\ }

86: \newcommand{\mMzero}{\mbox{$(m-M)_0$}\ }

87: \newcommand{\thetaspec}{\mbox{$\theta_{\mbox{\scriptsize spec}}$}}

88: \newcommand{\thetaphot}{\mbox{$\theta_{\mbox{\scriptsize phot}}$}}

89: \newcommand{\Smin}{\mbox{$\Sigma_{\mbox{\scriptsize min}}$}\,}

90: \newcommand{\Ab}{\mbox{$\mbox{A}_{\mbox{\scriptsize B}}$}\,}

91: \newcommand{\vrad}{\mbox{$\mbox{v}_{\mbox{\scriptsize rad}}$}\,}

92: \newcommand{\hubble}{\mbox{$H_{\mbox{\scriptsize 0}}\:$}}

93: \newcommand{\phase}[2]{\mbox{[{#1};{#2}]}\,}

94: \newcommand{\magdex}{\mbox{mag~dex$^{-1}\,$}}

95: 

96: 

97: \shorttitle{The angular size of the Cepheid $\ell$ Car}

98: \shortauthors{Kervella et al.}

99: 

100: \begin{document}

101: 

102: \title{The angular size of the Cepheid $\ell$ Car: a comparison

103: of the interferometric and surface brightness techniques}

104: 

105: \author{Pierre Kervella\altaffilmark{1,7},

106: Pascal Fouqu\'e\altaffilmark{1,2},

107: Jesper Storm\altaffilmark{3},

108: Wolfgang P. Gieren\altaffilmark{4},

109: David Bersier\altaffilmark{5},

110: Denis Mourard\altaffilmark{6},

111: Nicolas Nardetto\altaffilmark{6},

112: Vincent Coud\'e du Foresto\altaffilmark{7}

113: }

114: 

115: 

116: \altaffiltext{1}{European Southern Observatory, Casilla 19001,

117: Santiago 19, Chile}

118: 

119: \altaffiltext{2}{Observatoire Midi-Pyr\'en\'ees, UMR 5572, 14, avenue

120: Edouard Belin, F-31400 Toulouse, France}

121: 

122: \altaffiltext{3}{Astrophysikalisches Institut Potsdam, An der

123: Sternwarte 16, D-14482 Potsdam, Germany}

124: 

125: \altaffiltext{4}{Universidad de Concepci\'on, Departamento de

126: F\'{\i}sica, Casilla 160-C, Concepci\'on, Chile}

127: 

128: \altaffiltext{5}{Space Telescope Science Institute, 3700 San Martin

129: Drive, Baltimore, MD 21218, USA}

130: 

131: \altaffiltext{6}{GEMINI, UMR 6203, Observatoire de la C\^ote d'Azur,

132: Avenue Copernic, F-06130 Grasse, France}

133: 

134: \altaffiltext{7}{LESIA, Observatoire de Paris-Meudon, 5, place Jules

135: Janssen, F-92195 Meudon Cedex, France}

136: 

137: \email{pkervell@eso.org,

138: pfouque@eso.org,

139: jstorm@aip.de,

140: wgieren@coma.cfm.udec.cl,

141: bersier@stsci.edu,

142: Denis.Mourard@obs-azur.fr,

143: Nicolas.Nardetto@obs-azur.fr,

144: vincent.foresto@obspm.fr

145: }

146: 

147: 

148: \begin{abstract}

149: Recent interferometric observations of the brightest and angularly largest

150: classical Cepheid, $\ell$~Carinae, with ESO's VLT Interferometer (VLTI)

151: have resolved with high precision the variation of its angular diameter with phase.

152: We compare the measured angular diameter curve to

153: the one we derive by an application of the Baade-Wesselink type

154: infrared surface brightness technique, and find a

155: near-perfect agreement between the two curves.

156: The mean angular diameters of $\ell$~Car from the two techniques agree

157: very well within their total error bars (1.5\,\%), as do the derived distances (4\,\%).

158: This result is an indication that the calibration of the surface

159: brightness relations used in the distance determination of far away

160: Cepheids is not affected by large biases.

161: \end{abstract}

162: 

163: \keywords{Stars: distances -- Stars: fundamental parameters --

164: Stars: variables: Cepheids -- Stars: oscillations -- Techniques:

165: interferometry -- extragalactic distance scale}

166: 

167: 

168: \section{Introduction}

169: 

170: Cepheid variables are fundamental objects for the calibration of the

171: extragalactic distance scale. Distances of Cepheids can be derived in

172: at least two different ways: by using their observed mean magnitudes

173: and periods together with a period-luminosity relation, or by applying

174: a Baade-Wesselink (hereafter BW) type technique to determine their distances

175: and mean diameters from their observed variations in magnitude, color

176: and radial velocity.  This latter technique has been dramatically

177: improved by the introduction of the near-infrared surface brightness

178: method (hereafter IRSB) by \citet{Welch94}, and later by

179: \citet{FG97} who calibrated the relation between the $V$-band surface

180: brightness and near-infrared colors of Cepheids. For this purpose, they used

181: the observed interferometric angular diameters of a number of giants and

182: supergiants bracketting the Cepheid color range. This method has

183: been applied to a large number of Galactic Cepheid variables, for instance

184: by \citet{GFG97}, \citet{GFG98}, and \citet{Storm04}.

185: 

186: Applying the surface-brightness relation derived from stable

187: stars to Cepheids implicitly assumes that the relation

188: also applies to pulsating stars.

189: The validity of this assumption can now be addressed by comparing

190: direct interferometric measurements of the angular diameter variation of

191: a Cepheid to the one derived from the IRSB technique. It has recently

192: been shown by \citet{Kervella04}, hereafter K04, that the VLT Interferometer on

193: Paranal is now in a condition to not only determine accurate {\it mean}

194: angular diameters of nearby Cepheid variables, but to follow their

195: angular diameter {\it variations} with high precision.

196: Using the Palomar Testbed Interferometer,

197: \citet{lane00,lane02} resolved the pulsation of the Cepheids $\zeta$~Gem and

198: $\eta$~Aql as early as 2000, but the comparison we present in this

199: letter is the first where error bars on the derived distance and

200: linear diameter are directly comparable at a few percent level between

201: the interferometric and IRSB techniques.

202: 

203: The star that we will discuss in this letter, $\ell$~Car, is the brightest Cepheid

204: in the sky. Its long period of about 35.5 days implies a large

205: mean diameter, which together with its relatively short

206: distance makes it an ideal target for resolving its angular diameter

207: variations with high accuracy. In this paper, we will compare

208: the interferometrically determined angular diameter curve of $\ell$~Car

209: with that determined from the IRSB technique,

210: and we will demonstrate that the two sets of angular diameters are

211: in excellent agreement. Based on the available high-precision angular

212: diameter and radial velocity curves for this star, we will also derive

213: a revised value of its distance and mean radius.

214:  

215: Several authors \citep{Sasselov94, Marengo03, Marengo04}

216: have pointed out potential sources of systematic uncertainties

217: in the determination of Cepheid distances using the interferometric

218: BW method. In particular, imperfections in the numerical

219: modeling of Cepheid atmospheres could lead to biased estimates

220: of the limb darkening and projection factor. We will discuss the

221: magnitude of these uncertainties in the case of $\ell$~Car.

222: 

223: \section{Interferometric observations}

224: 

225: The interferometric observations of $\ell$~Car were obtained with the

226: VLT Interferometer \citep{Glindemann00}, using its

227: commissioning instrument VINCI \citep{Kervella00,Kervella03} and 0.35\,m test siderostats.

228: This instrument recombines the light from two telescopes in the infrared

229: $K$ band (2.0--2.4\,$\mu$m), at an effective wavelength of $2.18\,\mu$m.

230: A detailed description of the interferometric data recorded on $\ell$~Car

231: can be found in K04.

232: %All measurements

233: %except one have been obtained on the B4-M0 interferometric baseline,

234: %140\,m in ground length.  The observations of Julian date 2452755.617 were

235: %taken on the shorter E0-G1 baseline (66\,m ground length).

236: %The raw data have been reduced using the methods described by

237: %\citet{Kervella04b}, that make use of the wavelet

238: %transformation to estimate the visibility of the interferometric

239: %fringes with high precision. 

240: 

241: The limb darkening (LD) models used to derive the photospheric diameters from

242: the fringe visibilities were taken from \citet{Claret00}.

243: The correction introduced on the uniform disk (UD) interferometric measurements

244: by the limb darkening is small in the $K$ band: for $\ell$~Car, we determine

245: $k= \theta_{\rm UD}/\theta_{\rm LD} = 0.966$. Considering the magnitude of this

246: correction, a total systematic uncertainty of $\pm 1\,\%$ appears reasonable.

247: However, until the limb darkening of a sample of Cepheids has been measured

248: directly by interferometry, this value relies exclusively on numerical models of the

249: atmosphere. This is expected to be achieved in the next years using for instance

250: the longest baselines of the VLTI (up to 202\,m) and the shorter $J$ and $H$ infrared

251: bands accessible with the AMBER instrument \citep{Petrov00}.

252: 

253: The LD correction is changing slightly over the pulsation of the star, due

254: to the change in effective temperature, but \citet{Marengo03} have estimated the

255: amplitude of this variation to less than $0.3\,\%$ peak to peak in the $H$ band

256: (for the 10\,days period Cepheid $\zeta$~Gem). It is even lower in the $K$ band,

257: and averages out in terms of rms dispersion. As a consequence,

258: we have neglected this variation in the present study.

259: 

260: The limb-darkened angular diameter measurements

261: are listed in Table~\ref{table_angdiams_l_car}.

262: Two error bars are given for each point, corresponding respectively

263: to the statistical uncertainty (internal error) and to the systematic

264: error introduced by the uncertainties on the assumed angular diameters

265: of the calibrator stars (external error). The phases given in

266: Table~\ref{table_angdiams_l_car} are based on the new ephemeris derived

267: in Sect.\ref{sec.diameter}.

268: These measurements were obtained during the commissioning of the VLTI,

269: and part of them are affected by relatively large uncertainties (3-5\%)

270: due to instrumental problems.  However, the precision reached by VINCI

271: and the test siderostats on this baseline is of the order of 1\%

272: on the angular diameter, as demonstrated

273: around the maximum diameter phase.

274: 

275: \section{The infrared surface brightness technique}

276: 

277: The IRSB technique has been presented and

278: discussed in detail in \citet{FG97} (hereafter FG97).

279: In brief, the angular diameter curve of a

280: given Cepheid variable is derived from its $V$ light and $(V-K)$ color

281: curve, appropriately corrected for extinction.

282: It is then combined with its linear displacement, which is essentially

283: the integral of the radial velocity curve. 

284: A linear regression of pairs of angular diameters and linear

285: displacements, obtained at the same pulsation phases, yields both the

286: distance, and the mean radius of the star.

287: 

288: While there are several sources of systematic uncertainty in the method,

289: as discussed in \citet{GFG97}, one of its great advantages is

290: its strong insensitivity to the adopted reddening corrections, and to

291: the metallicity of the Cepheid \citep{Storm04}.

292: With excellent observational data at

293: hand, individual Cepheid distances and radii can be determined with an

294: accuracy of the order of 5\,\% {\em if} the adopted $K$-band surface

295: brightness-color relation is correct.

296: 

297: A first calibration of this relation coming directly from interferometrically

298: determined angular diameters of Cepheid variables was presented by

299: \citet{Nordgren2002} (herafter N02). They found a satisfactory agreement

300: with the FG97 calibration, within the combined 1\,$\sigma$ uncertainties

301: of both surface brightness-color calibrations.

302: Considering more closely the results from N02,

303: an even better agreement is found between the $F_V(V-K)$

304: relations \emph{before} the zero point if forced between the different colors.

305: Before this operation, N02 found the relation:

306: \begin{equation}

307: F_{V_0} = 3.956 \pm 0.011 - 0.134 \pm 0.005 (V - K)_0

308: \end{equation}

309: that translates, after forcing the zero point to the average

310: of the three selected colors, to the relation:

311: \begin{equation}

312: F_{V_0} = 3.941 \pm 0.004 - 0.125 \pm 0.005 (V - K)_0.

313: \end{equation}

314: On the other hand, FG97 obtain:

315: \begin{equation}

316: F_{V_0} = 3.947 \pm 0.003 - 0.131 \pm 0.003 (V - K)_0

317: \end{equation}

318: From this comparison, it appears that the slope initially

319: determined by N02 for $F_V(V-K)$ is significantly different both from their

320: final value and from the FG97 relation. This difference could cause a

321: bias due to the averaging of the multi-color zero points.

322: Though small in absolute value, such a bias is of particular importance 

323: for $\ell$~Car, due to its relatively large $(V-K)$ color.

324: 

325: Another argument in favor of the FG97 surface brightness relation

326: is that it relies on a sample of 11 Cepheids with periods of 4 to 39~days,

327: while the relations established by N02 were derived from the observations

328: of only 3 Cepheids with periods of 5 to 10~days.

329: Such short period Cepheids are significantly hotter

330: than $\ell$~Car ($P = 35.5$~days), and a local difference of the

331: slope of the IRSB relations cannot be excluded.

332: For these two reasons, we choose to retain the FG97 calibration

333: for our analysis of the Cepheid $\ell$~Car in the following section.

334: 

335: \section{Diameter and distance}

336: \label{sec.diameter}

337: 

338: \subsection{Angular diameter}

339: \label{subsec.angular}

340: 

341: We have combined the photometric data from \citet{Pel76} and

342: \citet{Bersier02} to construct the $V$-band light curve for

343: $\ell$~Car. The two data sets are spanning almost 30 years

344: and allow an improved determination of the period of this variable. We

345: find $P=35.54804$~days. The time of maximum $V$ light has been adopted

346: from \citet{Szabados89} who give a value of $T_0=2440736.230$

347: which is also in good agreement with the more recent data. The

348: resulting light curve is shown in Fig.\ref{fig.VK}.

349: The $K$ band light curve is based on the data from

350: \citet{LS92} and is also shown in Fig.\ref{fig.VK}. The $(V-K)$

351: color curve which is needed by the IRSB method has been

352: constructed on the basis of the observed $V$ band data and a Fourier fit to

353: the $K$-band data as described in \citet{Storm04}.

354: 

355: For the radial-velocity curve we have used the data from

356: \citet{Taylor97} and \citet{Bersier02}.  Using the

357: new ephemeris from above we detected a slight offset of 1.5~\kms

358: between the two datasets. We choose to shift the

359: \citet{Taylor97} dataset by $-1.5$~\kms to bring all the data on

360: the well established CORAVEL system of \citet{Bersier02}.  We

361: note that the exact radial velocity zero point is irrelevant as the

362: method makes use of \emph{relative} velocities. The combined

363: radial velocity data are displayed in Fig.\ref{fig.VK}.

364: 

365: 

366: The application of the IRSB method has

367: followed the procedure described in \citet{Storm04}.

368: We have adopted the same reddening law with

369: $R_V=3.30$ and $R_K=0.30$, a reddening of $E(B-V)=0.17$ 

370: \citep{Fernie90}, and a projection factor, $p$, from radial to

371: pulsational velocity of $p=1.39-0.03\log P=1.343$ \citep{Hindsley86,GBM93}.

372: As discussed by \citet{Storm04} we only consider the

373: points in the phase interval from $0.0$ to $0.8$ (phase zero is

374: defined by the $V$ band maximum light).

375: We have applied a small phase shift of $-0.025$ to the radial velocity

376: data to bring the photometric and radial velocity based angular

377: diameters into agreement.  We note that a similar phase shift can be

378: achieved by lowering the systemic velocity by $1.5$~\kms.

379: 

380: The angular diameter curve obtained from the photometry has been

381: plotted in Fig.\ref{fig.svb}, together with the linear displacement

382: curve.  The photometric and interferometric diameter curves are

383: directly compared in Fig.~\ref{fig.LDfit}, where they are plotted as a

384: function of phase.

385: With these data we can compute the average

386: angular diameters obtained from each technique. 

387: For the IRSB, we find an average limb darkened angular diameter

388: $\overline{\theta_{\rm LD}} = 2.974 \pm 0.046$ milliarcsecond, and for

389: the interferometric measurements we find $\overline{\theta_{\rm LD}} =

390: 2.992 \pm 0.012$ milliarcsecond. The

391: %(see Table~\ref{table_l_car})

392: agreement between these two values is strikingly good.  This is a

393: serious indication that the calibration of the surface

394: brightness-color relation (FG97), based on non-pulsating giant stars,

395: does apply to Cepheids.

396: 

397: \subsection{Distance}

398: \label{subsec.distance}

399: 

400: The surface brightness method yields a

401: %distance modulus of $(m-M)_0 = 8.74\pm0.02$, corresponding to a

402: distance of $560 \pm 6$~pc, and a mean radius of $R=179 \pm 2$~\Rsolar.

403: The corresponding mean absolute $V$ magnitude is

404: $M_V=-5.57 \pm 0.02$~mag and the distance modulus is $(m-M)_0 = 8.74

405: \pm 0.05$.  The error estimates are all intrinsic 1$\sigma$ random

406: errors.  In addition to these random errors, a systematic error of the

407: order of 4\% should be taken into account, as discussed by \citet{GFG97}.

408: The final IRSB values are thus $d = 560 \pm 23$~pc,

409: and $R = 179 \pm 7$~\Rsolar.

410: Compared to \citet{Storm04} we find a significantly

411: (0.24~mag) shorter distance modulus for $\ell$~Car. This can be

412: explained by the use in the present Letter of the new and superior

413: radial velocity data from \citet{Taylor97} and \citet{Bersier02}.

414: 

415: K04 found $d = 603^{+24}_{-19}$ pc, using the interferometric angular diameters

416: and a subset of the radial velocity data used here. To make the

417: comparison more relevant, we determined the distance and radius using

418: the same data (interferometric diameters from

419: Table~\ref{table_angdiams_l_car} and radial velocity from

420: \citet{Taylor97} and \citet{Bersier02} -- see above),

421: the same ephemeris, and the same

422: projection factor (see Sect~\ref{subsec.angular}). Using the method of

423: K04, we find a distance $d = 566^{+24}_{-19}$, and a linear radius

424: $R = 182^{+8}_{-7}$~\Rsolar. This is in excellent agreement with the values

425: obtained from the IRSB method.

426: % (Table~\ref{table_l_car}).

427: 

428: This 6\,\% difference in the distances based on interferometric diameters

429: (603\,pc for K04 versus 566\,pc here) has two major causes. First, the

430: $p$-factor used in the present paper is $\sim 1.3$\% smaller than in K04.

431: The choice of the reference used for the $p$-factor has currently an impact

432: of a few percents on its value. This indicates that the average value of the

433: $p$-factor for a given Cepheid is currently uncertain by at least a similar

434: amount, and this systematic error translates linearly to the distance

435: determination.

436: 

437: Secondly, the use of a different -- and superior -- data set for the

438: radial velocity makes the radius curve different from K04. In

439: particular the amplitude is smaller here than in K04 by $\sim 3$\%. This

440: is likely due to the more complete phase coverage that we have here, and

441: possibly also to a different choice of spectral lines to estimate

442: the radial velocity. This amplitude difference translates linearly on the

443: distance through the BW method.

444: 

445: \section{Conclusion}

446: 

447: The main point of our paper is to show that with a

448: consistent treatment of the data, the internal accuracy of both methods

449: (IRSB or interferometry) is extremely good: the angular diameter variation

450: observed using the VLTI agrees very well with that derived from the

451: $F_V(V-K)$ version of the IRSB technique as calibrated by FG97.

452: For all the interferometric measurements, the corresponding IRSB angular

453: diameter at the same phase lies within the combined 1$\sigma$ error

454: bars of the two measurements (Fig.~\ref{fig.LDfit}).

455: Even more importantly, the mean angular diameter of the Cepheid as

456: derived from both independent sets of angular diameter determination

457: are in excellent agreement, within a few percents.

458: 

459: Unfortunately, this is not equivalent to say that the Cepheid

460: distance scale is calibrated to a 1\% accuracy. We have drawn

461: attention to remaining sources of systematic errors that can

462: affect Cepheid radii and distances up to several percents.

463: As an illustration of these sources, K04 obtain a distance $d= 603^{+24}_{-19}$\,pc

464: for $\ell$~Car, while we obtain $d = 566^{+24}_{-19}$\,pc from the same

465: interferometric data.

466: 

467: We have already shown that most of the 6\,\% difference

468: (equivalent to $1.3\,\sigma$) can be explained by the use of different radial

469: velocity data and projection factor. Another thing to consider is the phase interval

470: used. K04 used measurements over the whole

471: pulsation cycle whereas in the IRSB technique, one

472: avoids the phase interval 0.8--1 (Fig.~\ref{fig.svb}).

473: During that phase interval, that corresponds to the rebound of

474: the atmosphere around the minimum radius, energetic shock waves are

475: created. As discussed by \citet{Sabbey95}, they produce asymmetric line

476: profiles in the Cepheid spectrum. Recent modeling using a self-consistent dynamical

477: approach also show that the $\tau=1$ photosphere may not be comoving with

478: the atmosphere of the Cepheid during its pulsation, at the 1\,\% level \citep{Nardetto04}.

479: Such an effect would impact the $p$-factor, modify the shape of

480: the radial velocity curve, and thus bias the amplitude of the radius variation,

481: possibly up to a level of a few percents. As the BW method (either its classical or

482: interferometric versions) relies linearly on this amplitude, a bias at this level

483: presently cannot be excluded.

484: 

485: The interferometric BW method is currently limited to

486: distances of 1--2 \,kpc, due to the limited length of the available

487: baselines. The IRSB technique on the other hand 

488: can reach extra-galactic Cepheids as already demonstrated by

489: \citet[for the LMC]{gieren00} and by \citet[for the SMC]{Storm04}.

490: Using high precision interferometric measurements of $\ell$~Car and

491: other Cepheids, it will be possible to calibrate the IRSB

492: method down to the level of a few percents.

493: From the present comparison, we already see that

494: this fundamental calibration will be very similar to the calibration

495: found by FG97 and \citet{Nordgren2002}.

496: 

497: 

498: \acknowledgments

499: 

500: WPG acknowledges support for this work from the chilean FONDAP Center

501: for Astrophysics 15010003.

502: 

503: \newpage

504: 

505: \begin{thebibliography}{}

506: \bibitem[Bersier(2002)]{Bersier02}

507: Bersier, D. 2002, \apjs, 140, 465

508: \bibitem[Claret(2000)]{Claret00}

509: Claret, A. 2000, \aap, 363, 1081

510: \bibitem[Fernie(1990)]{Fernie90}

511: Fernie, J.D. 1990, \apjs, 72, 153

512: \bibitem[Fouqu\'e \& Gieren(1997)]{FG97}

513: Fouqu\'e, P., \& Gieren, W.P. 1997, \aap, 320, 799 (FG97)

514: \bibitem[Gieren et al.(1997)]{GFG97}

515: Gieren, W.P., Fouqu\'e, P., \& G\'omez, M. 1997, \apj, 488, 74 (GFG97)

516: \bibitem[Gieren et al.(1998)]{GFG98}

517: Gieren, W.P., Fouqu\'e, P., \& G\'omez, M. 1998,  \apj, 496, 17

518: \bibitem[Gieren et al.(1993)]{GBM93}

519: Gieren, W.P., Barnes, T.G., \& Moffett, T.J. 1993,  \apj, 418, 135

520: \bibitem[Gieren et al.(2000)]{gieren00}

521: Gieren, W.P., Storm, J., Fouqu\'e, P., Mennickent, R. E., \& G\'omez, M. 2000,  \apjl, 533, L107

522: \bibitem[Glindemann et al.(2000)]{Glindemann00}

523: Glindemann, A., Abuter, R., Carbognani, F., et al. 2000, \procspie, 4006, 2

524: \bibitem[Hindsley \& Bell.(1986)]{Hindsley86}

525: Hindsley, R. B., \& Bell, R. A. 1986, \pasp, 98, 881

526: \bibitem[Kervella et al.(2000)]{Kervella00}

527: Kervella, P., Coud\'e du Foresto, V., Glindemann, A., \& Hofmann, R. 2000, \procspie, 4006, 31

528: \bibitem[Kervella et al.(2003)]{Kervella03}

529: Kervella, P., Gitton, Ph., S\'egransan, D., et al. 2003, \procspie, 4838, 858

530: \bibitem[Kervella et al.(2004)]{Kervella04}

531: Kervella, P., Nardetto, N., Bersier, D., et al. 2004, \aap, in press, astro-ph/0311525 (K04)

532: \bibitem[Lane et al.(2000)]{lane00} Lane, B.~F., Kuchner, M.~J., Boden, A.~F.,

533: Creech-Eakman, M. , Kulkarni, S.~R. 2000, Nature, 407, 485

534: \bibitem[Lane et al.(2002)]{lane02}

535: Lane, B. F., Creech-Eakman, M. J., \& Nordgren, T. E. 2002, \apj, 573, 330

536: \bibitem[Laney \& Stobie(1992)]{LS92}

537: Laney, C. D., \& Stobie, R.S 1992, A\&A Suppl. Ser., 93, 93

538: \bibitem[Marengo et al.(2003)]{Marengo03}

539: Marengo, M., Karovska, M., Sasselov, D. D., et al. 2003, \apj, 589, 968

540: \bibitem[Marengo et al.(2004)]{Marengo04}

541: Marengo, M., Karovska, M., Sasselov, D. D., \& Sanchez, M. 2004, \apj, 603, in press, astro-ph/0312349

542: \bibitem[Nardetto et al.(in preparation)]{Nardetto04}

543: Nardetto, N., et al. 2004, in preparation

544: \bibitem[Nordgren et al.(2002)]{Nordgren2002}

545: Nordgren, T. E., Lane, B. F., Hindsley, R. B., \& Kervella, P. 2002, \apj, 123, 3380 (N02)

546: \bibitem[Pel(1976)]{Pel76}

547: Pel, J. W. 1976, A\&A Suppl. Ser., 24, 413

548: \bibitem[Petrov et al.(2000)]{Petrov00}

549: Petrov, R., Malbet, F., Richichi, A., et al. 2000, \procspie, 4006, 68

550: \bibitem[Sabbey et al.(1995)]{Sabbey95}

551: Sabbey, C.N., Sasselov, D.D., Fieldus, M.S., et al. 1995, ApJ, 446, 250

552: \bibitem[Sasselov \& Karovska(1994)]{Sasselov94}

553: Sasselov, D. D. \& Karovska M. 1994, \apj, 432, 367

554: \bibitem[Storm et al.(2004)]{Storm04}

555: Storm, J., Carney, B.W., Gieren, W.P., Fouqu\'e, P., Latham, D.W., \& Fry, A.M. 2004, \aap, in press, astro-ph/0401211

556: \bibitem[Szabados(1989)]{Szabados89}

557: Szabados, L. 1989, Comm. Konkoly Observatory, 94, 1

558: \bibitem[Taylor et al.(1997)]{Taylor97}

559: Taylor, M.M., Albrow, M.D., Booth, A.J., \& Cottrell, P.L. 1997, \mnras, 292, 662

560: \bibitem[Welch(1994)]{Welch94}

561: Welch, D.L. 1994, \aj, 108, 1421

562: \end{thebibliography}

563: 

564: 

565: %___________________Table of angular diameter measurements

566: \begin{table}

567: \caption{Angular diameter measurements of $\ell$~Car. The statistical

568: and systematic calibration uncertainties are mentioned separately in brackets.

569: \label{table_angdiams_l_car}}

570: \begin{tabular}{ccc}

571: \tableline\tableline

572: \noalign{\smallskip}

573: Julian Date & Phase & $\theta_{\rm LD}$ (mas) \\

574: \tableline

575: \noalign{\smallskip}

576:  2452453.498 & 0.618 & $3.054 \pm 0.113\ _{[0.041, 0.105]}$\\

577:  2452739.564 & 0.665 & $2.891 \pm 0.087\ _{[0.076, 0.043]}$\\

578:  2452740.569 & 0.693 & $2.989 \pm 0.047\ _{[0.018, 0.044]}$\\

579:  2452741.717 & 0.726 & $2.993 \pm 0.039\ _{[0.026, 0.029]}$\\

580:  2452742.712 & 0.754 & $2.899 \pm 0.056\ _{[0.035, 0.043]}$\\

581:  2452743.698 & 0.781 & $2.758 \pm 0.076\ _{[0.074, 0.016]}$\\

582:  2452744.634 & 0.808 & $2.794 \pm 0.035\ _{[0.032, 0.013]}$\\

583:  2452745.629 & 0.836 & $2.675 \pm 0.098\ _{[0.097, 0.017]}$\\

584:  2452746.620 & 0.864 & $2.775 \pm 0.046\ _{[0.023, 0.040]}$\\

585:  2452747.599 & 0.891 & $2.699 \pm 0.129\ _{[0.127, 0.026]}$\\

586:  2452749.576 & 0.947 & $2.645 \pm 0.078\ _{[0.077, 0.012]}$\\

587:  2452751.579 & 0.003 & $2.753 \pm 0.033\ _{[0.028, 0.017]}$\\

588:  2452755.617 & 0.117 & $2.970 \pm 0.113\ _{[0.113, 0.013]}$\\

589:  2452763.555 & 0.340 & $3.194 \pm 0.034\ _{[0.009, 0.033]}$\\

590:  2452765.555 & 0.396 & $3.212 \pm 0.034\ _{[0.011, 0.033]}$\\

591:  2452766.550 & 0.424 & $3.210 \pm 0.035\ _{[0.011, 0.033]}$\\

592:  2452768.566 & 0.481 & $3.188 \pm 0.037\ _{[0.011, 0.035]}$\\

593:  2452769.575 & 0.509 & $3.189 \pm 0.022\ _{[0.018, 0.012]}$\\

594:  2452770.535 & 0.536 & $3.160 \pm 0.022\ _{[0.020, 0.009]}$\\

595:  2452771.528 & 0.564 & $3.136 \pm 0.020\ _{[0.017, 0.010]}$\\

596:  2452786.620 & 0.989 & $2.727 \pm 0.064\ _{[0.012, 0.063]}$\\

597: \noalign{\smallskip}

598: \tableline

599: \end{tabular}

600: \end{table}

601: 

602: %%___________________Table

603: %\begin{table}

604: %\caption{Average angular diameter, distance and radius of $\ell$~Car.

605: %\label{table_l_car}}

606: %\begin{tabular}{lccc}

607: %\tableline\tableline

608: %\noalign{\smallskip}

609: %Method & $\overline{\theta_{\rm LD}}$ (mas) & $d$ (pc) & $R$ (\Rsolar) \\

610: %\tableline

611: %\noalign{\smallskip}

612: %Photometry & $2.974 \pm 0.046$ & $560 \pm 23$ & $179 \pm 8$ \\

613: %Interferometry & $2.992 \pm 0.012$ & $566^{+24}_{-19}$ & $182^{+8}_{-7}$ \\

614: %\noalign{\smallskip}

615: %\tableline

616: %\end{tabular}

617: %\end{table}

618: 

619: %_______________________ New page

620: \newpage

621: 

622: %_______________________ Figure

623: \begin{figure}[htp]

624: \plotone{f1.ps}

625: \caption{\label{fig.VK}

626: Radial velocity curve of $\ell$~Car (upper panel)

627: using data from \citet{Taylor97} shifted by $-1.5$~\kms

628: (circles) and from \citet{Bersier02} (triangles).

629: The $K$ band photometric measurements (middle panel) were taken from

630: \citet{LS92}. We have relied on \citet{Pel76} (circles) and

631: \citet{Bersier02} (triangles) for the $V$ band data.}

632: \end{figure}

633: 

634: %_______________________ Figure

635: \begin{figure}[htp]

636: \plotone{f2.ps}

637: \caption{\label{fig.svb} Photometric angular diameters plotted against

638: phase for our best fitting distance.  The solid curve represents the

639: integrated radial velocity curve of $\ell$~Car for the adopted

640: distance.}

641: \end{figure}

642: 

643: 

644: %_______________________ Figure

645: \begin{figure}[htp]

646: \plotone{f3.ps}

647: \caption{\label{fig.LDfit} The interferometrically determined angular

648: diameters, plotted against phase (solid dots) with the angular

649: diameters derived with the infrared surface brightness method overplotted

650: (crosses). In the upper right corner a typical error bar 

651: for the surface brightness method data is shown.}

652: \end{figure}

653: 

654: 

655: \end{document}

656: