A proper-motion survey for white dwarfs with the Wide Field Planetary Camera 2

We have performed a search for halo white dwarfs as high proper motion objects in a second-epoch Wide Field Planetary Camera 2 image of the Groth-Westphal strip. The survey covers 74.8 arcmin(2) and is complete to V similar to 26.5. We identify 24 high proper motion objects with mu > 0.014 yr(-1). Five of these high proper motion objects are identified as strong white dwarf candidates on the basis of their position in a reduced proper motion diagram. We also identify two marginal candidates whose photometric errors place them within similar to1 sigma of the white dwarf region of the reduced proper motion diagram. We create a model of the Milky Way thin disk, thick disk, and stellar halo and find that this sample of white dwarfs is clearly an excess above the less than or equal to2 detections expected from these known stellar populations. The origin of the excess signal is less clear. Possibly, the excess cannot be explained without invoking a fourth galactic component: a white dwarf dark halo. Previous work of this nature has separated white dwarf samples into various galactic components based on kinematics; distances, and thus velocities, are unavailable for a sample this faint. Therefore, we present a statistical separation of our sample into the four components and estimate the corresponding local white dwarf densities using only the directly observable variables, V, (V-I), and mu. For all Galactic models explored, our five white dwarf sample separates into about three disk white dwarfs and two halo white dwarfs. However, the further subdivision into the thin and thick disk and the stellar and dark halo, and the subsequent calculation of the local densities, are sensitive to the input parameters of our model for each Galactic component. Using the lowest mean mass model for the dark halo and the five white dwarf sample, we find n(0,thin disk) = 2.4(-0.6)(0.7) x 10(-2) pc, n(0, thick disk) = 0.0(+7.6) x 10(-4) pc(-3), n(0,) (stellar halo) = 0.0(+7.7) x 10(-5) pc(-3), and n(0, dark halo) = 1.0(-0.4)(+0.4) x 10(-3) pc(-3). This implies a 7% white dwarf halo and 6 times the canonical value for the thin disk white dwarf density (at marginal statistical significance), but possible systematic errors due to uncertainty in the model parameters likely dominate these statistical error bars. The white dwarf halo can be reduced to similar to1.5% of the halo dark matter by changing the initial mass function slightly. The local thin disk white dwarf density in our solution can be made consistent with the canonical value by assuming a larger thin disk scale height of 500 pc.