We have used the Hipparcos parallaxes of nearby Galactic Cepheids to explore corrections to the multiwavelength Period-Luminosity relations for LMC Cepheids. The latter are based on an LMC data set scaled to a true distance modulus of 18.50 mag and an adopted foreground reddening of E (B-V) = 0.10 mag. Although the current uncertainties in the parallaxes are large and still dependent upon the specific subsets of the Cepheids chosen for the comparison, the agreement is good, indicating that to within ± 0.14 mag (or, 7% in distance) the previously adopted zero point is substantially correct. Based on different subsamples of data either having BV, BVJK, BVIJK or BVIJHK photometry, LMC moduli, ranging from 18.44 to 18.57 mag are derived. These results, summarized in Table 2, differ from the value of 18.70 mag of FC97, which are based solely on the reddening-corrected V photometry of Caldwell & Laney (1991), externally adjusted for metallicity. The Hipparcos data alone do not allow us to discriminate between metallicity effects and the physically distinct possibility of added reddening to the LMC.
| Filters | No. Stars | E (B-V) ±  |
µ LMC ± |
|---|---|---|---|
| BV | 19 | 0.17 ± . . . | 18.44 ± 0.35 |
| BVJK | 13 | 0.14 ± 0.08 | 18.50 ± 0.13 |
| BVIJK | 10 | 0.16 ± 0.07 | 18.53 ± 0.14 |
| BVIJHK | 7 | 0.17 ± 0.07 | 18.57 ± 0.11 |
| VcJHK | 8 | 0.16 ± 0.11 | 18.57 ± 0.11 |
To alleviate the ambiguity posed by the need to simultaneously solve
both for reddening and metallicity effects on the Cepheid distances we
are currently deriving OB-star reddenings along the individual lines
of sight to several dozen LMC Cepheids. This will allow us to
decouple the reddening determinations from metallicity effects, and go
beyond the use of a single mean (foreground + internal) reddening
for the LMC calibrating Cepheid sample. Preliminary
reductions
indicate that the variance from field to field is large (ranging from
E (B-V) = 0.00 up to 0.40 mag) while still indicating that an average
value of
We close by noting that at least three other very recent determinations of the true modulus to the LMC fall on either side of the value 18.50 mag adopted by MF91 in setting a zero point for the Cepheid distance scale. Both Reid (1997) and Gratton et al. (1997) derive large LMC moduli (18.65 ± 0.10, and 18.63 ± 0.06 mag, respectively) using Hipparcos-based calibrations of the Galactic globular cluster and RR Lyrae distance scale. On the other hand, Gould & Uza (1997) have re-analyzed the SN 1987A supernova ``light echo'' and derive an upper limit of µ LMC < 18.37 ± 0.04 mag for the LMC true distance modulus; although they note that if the ring is slightly elliptical (b / a ~ 0.95) this upper limit increases to < 18.44 ± 0.05 mag. A value of 18.56 ± 0.05 mag has been derived by Panagia et al. (1996) from the same data. Until these differences are fully understood and resolved, and given the remaining uncertainties in the Hipparcos Cepheid parallax data, we prefer to adopt a true distance modulus of 18.50 mag for the LMC, but now bounded by an uncertainty of ± 0.15 mag, defined to fully encompass the above range of recently published values. This value is consistent with other estimated distances to the LMC based on a wide variety of methods (for a comprehensive modern review see Westerlund 1997). Viewed in that perspective, the Hipparcos data confirm the Cepheid distance scale at better than the ± 10% level (95% confidence).