On the parallax of WD 0346+246: a halo white dwarf candidate

N. C. Hambly 2 w S. J. Smartt 1 S. T. Hodgkin 0 R. F. Jameson 0 S. N. Kemp 5 W. R. J. Rolleston 4 I. A. Steele 3 0 Astronomy Group, Department of Physics and Astronomy, Leicester University, University Road , Leicester LE1 7RH 1 Isaac Newton Group of Telescopes, Apartado de Correos 321, 38770 Santa Cruz de La Palma, La Palma, Islas Canarias, Spain 2 Institute for Astronomy, University of Edinburgh , Blackford Hill, Edinburgh EH9 3HJ 3 Astrophysics Research Institute, Liverpool John Moores University , Twelve Quays House, Egerton Wharf, Birkenhead L41 1LD 4 Department of Pure and Applied Physics, Queen's University of Belfast , Belfast BT7 1NN 5 Instituto de Astrofsica de Canarias , 38200 La Laguna, Tenerife, Islas Canarias, Spain A B S T R A C T We present a parallax measurement for the very cool degenerate WD 03461246, the serendipitous discovery of which was reported by Hambly et al. We find an absolute parallax of 36 ^ 5 mas, yielding a distance estimate of 28 ^ 4 pc. The resulting absolute visual magnitude of the object is MV 16:8 ^ 0:3, making it the second-lowest luminosity white dwarf currently known. We use the distance estimate and measured proper motion to show that the object has kinematics consistent with membership of the Galactic halo. WD 03461246 is therefore by far the coolest and least luminous of only a handful of plausible halo white dwarf candidates. As such, the object has relevance to the ongoing debate concerning the results of microlensing experiments and the nature of any baryonic dark matter component to the Galactic halo residing in stellar remnants. I N T R O D U C T I O N In a previous paper, we reported the serendipitous discovery of a faint, high proper motion star (Hambly, Smartt & Hodgkin 1997, hereafter HSH97). While there was no doubt as to the nature of the object spectroscopy unequivocally demonstrated it to be a cool white dwarf (hereafter CWD) there was some uncertainty as to its evolutionary status. No definitive statement could be made as to the kinematic population of the star because its space velocity depended on the unknown distance. For example, an intrinsic luminosity could not be assumed for the object owing to its unknown mass and therefore unknown radius. As discussed in HSH97, such CWDs belonging to the Galactic disc provide a useful lower limit to the age of that population (e.g. Knox, Hawkins & Hambly 1999 and references therein). On the other hand, a significant number of CWDs found belonging to the population II could go some way to resolving the current debate concerning the nature of the dark lensing bodies in the Galactic halo (e.g. Isern et al. 1998 and references therein). In this short paper, we present a parallax measurement for the star that resolves the uncertainty in distance and indicates that it is probably a degenerate belonging to the halo population. Section 2 describes the observations and data reduction; Section 3 describes In order to measure the parallax of WD 03461246 we observed the star, using the TEK4 charge-coupled device (CCD) camera and 1.0-m Jacobus Kapteyn Telescope on the island of La Palma, at nine separate epochs. Eight of these were spaced over the six months of the most recent observing season (1998/1999) starting and ending at extreme parallax factor. Table 1 provides a journal of the observations. Where possible, three separate I-band exposures were taken per observing session, at hour angles within ^30 min to minimize differential colour refraction effects (DCR e.g. Monet et al. 1992). On nights considered to be photometric, standard stars from the list of Landolt (1992) were observed. The CCD data were reduced following standard procedures within the iraf environment. In all cases, nightly mean bias frames showed no large or small scale non-uniformities so all frames were debiased using a straight-line fit to the overscan region; sky flats were used to flat-field the data. Aperture photometry and object centroiding were achieved using tasks from the iraf digiphot.daophot package. Standard aperture photometry procedures were followed throughout the photometric reductions (e.g. Massey & Davis 1992). Object centres were determined using intensity-weighted centroiding of marginal sums in x and y over pixels in each stellar image at a threshold of 3s above local sky (note that although full profile fitting is optimal in terms of centroiding accuracy, for well-exposed images the gain over intensity-weighted centroids is not expected to be large e.g. Irwin 1985). In order to measure the parallax of the target, each frame was mapped on to a common coordinate system defined by a set of 35 reference stars. These were chosen so as to have similar colour to the target, again to minimize DCR effects (note that using I-band frames as opposed to V for the astrometry also reduces these effects). Once on the same xy coordinate system, a single global astrometric model, defined by Schmidt plate secondary standards (Hambly et al. 1999) tied into the Tycho-ACT reference frame, Date 1997 February 16 1998 August 16 1998 August 24 1998 September 4 1998 October 21 1998 November 8 1999 January 31 1999 February 15 1999 February 19 No. of frames (astrometry) 1.076 1.160 1.1331.204 1.0021.003 1.0021.007 1.0021.004 1.0021.012 1.0161.036 1.0411.066 FWHM of images/00 was used to put each set of measurements on to the celestial coordinate frame. R E S U LT S Fig. 1 shows plots of the position of the object with respect to a reference point (defined below) as a function of Modified Julian Date of observation. The error bars on each point were computed as the quadrature sum of the centroiding error plus the error in transforming each individual set of measurements on to the common frame; for clarity of presentation, the curves are also shown with the gross linear component (i.e. proper motion) removed leaving the parallax signature. The astrometric parameters were found using a standard linear least-squares fit of the data to the system of normal equations defined by a classical parallax treatment (e.g. Green 1985) as follows: a0 1 ma t 1 p sec d X t sin a 2 Y t cos a where X(t)Y(t)Z(t) define the barycentric position vector of the Earth at any given time t. The positional astronomy library SLALIB (Wallace 1998) was used to provide these values as a function of MJD. The five parameters (a 0, d 0, m a , m d , p ) were fitted, and the residuals of the data about the fit are also shown in Fig. 1. The reduced chi-squared values for the fits are xn2 a; d 1:57; 1:14 . In Table 2 we give the astrometric results, where the errors on the proper motion components and parallax were obtained by propagating the measurement errors through the equations of the least-squares fit in quadrature. The relative parallax measured via the reference stars was converted to an absolute value using the photometric parallax relation of Stobie, Ishida & Peacock (1989) applied to the 35 reference stars. This yielded a med (...truncated)