Studies of HD 65949

1 March 2010: A complete abundance study of HD 65949 has now been accepted for publication in Monthly Notices of the RAS. See the astro-ph preprint arXiv:1002.2945. Some details are given here

Wavelengths and tentative identifications from 3285 - 4510[A]> for UVES spectrum obtained 8 Aug 2008.

Cowley, Hubrig, Gonzalez, and Nunez (Astron. Astrophys. Letters, 455, L21, 2006, Paper I) discussed the very unusual spectrum of HD 65949, a mid-B star in the cluster NGC 2516. Click for measured wavelengths and identifications.

We have relied most strongly on the profiles of the low Balmer lines to give the surface gravity. For surface temperatures between 12,600K and 13,600K, these profiles are well fit with Log(g) = 4.0. We adopt the effective temperature that gives equal abundances from Fe I, Fe II, and Fe III: T_eff = 12,600K.

The CDS (Strasbourg) gives UBV, Strömgren, and H-beta measurements. Several implementations give T_eff and either Log(g) or a stellar radius. These give effective temperatures between 13,200 and 13,600K. In our original paper, we adopted the higher value. Because the iron equilibrium indicated a temperature 1000K lower, we explored several other recipes for extracting T_eff from photometry. A well-known method was discussed by Johnson and Morgan (ApJ, 117, 313, 1953), known as the Q-method. We used

Q = (U-B) - 0.72 (B-V). Johnson and Morgan calibrated spectral types of O and B stars as a function of Q, which is independent of the reddening. The coefficient 0.72 is E(U-B)/E(B-V) if the reddening is "standard." We obtained a spectral type of B6.59 from Q = -0.35, and then used Table 15.7 of Astrophysical Quantities IV to find T_eff = 12,189K. For a B6.59 star, we expect (B-V) = -0.14. The observed value is -0.02. This leads to E(B-V) = 0.12, in good agreement with values found for NGC 2516.

It is presently unclear why the photometric temperatures are so different from one another. One possibility is the influence of a companion on the color.