1999ApJ...511..422M

The p-mode eigenfrequency spacings of 12 rapidly oscillating Ap (roAp) stars provide estimates of their radii. By combining these with independent estimates of their effective temperatures via Hβ photometry, Martinez predicted luminosities for all but one of this sample. Hipparcos parallaxes, and hence luminosities, are now available for the entire sample. While we obtain general agreement between the two sets of measurements, Hipparcos finds systematically larger parallaxes than predicted by asteroseismology. If the Hipparcos parallaxes and the interpretation of the p-mode spacings are both correct, this suggests the roAp stars are systematically cooler than expected from the Hβ calibration. Other explanations (e.g., systematic overestimates of the radii if Ap stars actually have globally metal-poor compositions and magnetic perturbations of the pulsational frequencies) are unlikely to account for the discrepancy. Lower temperatures for the roAp stars would shift many of them beyond the classical instability strip in the H-R diagram, implying that the He II ionization mechanism cannot be responsible for their pulsation. However, it is essential to ensure that both the Hipparcos parallaxes and the asteroseismic models do not suffer some systematic bias. We identify one case (HD 166473) where the parallax in the Tycho Catalogue must be wrong, giving a luminosity of ∼0.4 L_☉ for a late A star, and another (α Cir) where Hipparcos predicts a temperature 1000 K cooler than a detailed model of the stellar spectrum.