Astron. Astrophys. 323, 429-441 (1997)

5. Conclusions

Our spectroscopic survey at medium temporal and high spectral resolution of about a dozen bright F- and early G-type main-sequence and giant stars with broad spectral lines has confirmed that Doppler 'mapping' techniques can be quite effective in detecting pulsational surface velocity inhomogeneities in broad-lined stars also at relatively low effective temperatures. Nevertheless, this method quickly reaches fundamental limits for stars fainter than, say, 5th magnitude, with 4-m class telescopes, for periods shorter than 15-20 minutes, and in detailed time series analyses of multi-mode pulsators.

For most of our targets, we could only derive upper limits of 2-6 km/s for the amplitudes of sectoral p -modes with azimuthal indices between 6 and 16. For g -modes with larger horizontal-to-vertical amplitude ratio these numbers would be smaller but the associated physical velocities would go up. The small absolute numbers and the low detection rate do not permit a spectroscopic variability map of the region in question of the HR diagram to be derived. If the incidence of line profile variability (2 detections in 6 stars with suitable observations) is independent of the rotation rate, a simple extrapolation of our results without an attempt to correct for incompleteness suggests that in the HRD between the Scuti instability strip and the sun the fraction of nonradially pulsating stars with substantial amplitudes is up to 30%. Because one of the two stars ( Aql) still falls just within the Scuti instability strip, 15% is a more conservative estimate; but the very small sample size makes any such number extremely uncertain anyway.

Since the original aim of our experiment was to search for p -mode nonradial pulsation, the observations do not adequately sample the very long time scales of the variations of the line profile variability actually found in Ara and Aql. At the present level of observational incompleteness, the data are most easily compatible with nonradial pulsation. Only in the context of g -mode pulsation do the apparent symmetry in equivalent width of quasi-absorption and -emission features, their migration over the full width of the spectral lines, and a large amplitude in the line wings not require additional ad hoc assumptions for their explanation. At = 0.311, Ara lies halfway between the cool edge of the Scuti instability strip and the sun and would therefore be one of the spectroscopically identified nonradially pulsating stars closest to the sun in the HR diagram.

The only other nonradially pulsating stars known in this region of the HR diagramm are the Dor variables and the sun. While g -modes have not yet been unambiguously identified in the sun, long-period g -modes seem to be the defining characteristic of the Dor variables (Aerts & Krisciunas 1996; Balona et al. 1996). This makes it tempting to associate the line profile-variable stars found by this study with Dor stars. In this case, the higher order modes possibly present in Ara and Aql would be an extension of the present profile of this class of variables to larger m values. However, this tentative interpretation needs to gain in substance from a more detailed study. Similarly, a larger sample is needed to test whether it is significant that Ara and Aql are among the most 'normal' stars observed by us (assuming that binarity is not a restriction of normality).

For the more rapidly rotating program stars our observations cover at least one rotation period, albeit with poor sampling. The apparent constancy of the line profiles shows therefore that the chromospheric X-ray emission detected in many of the targets is not rooted in or above prominent photospheric patches (unless the flux originates from a cool companion indications for which were found in one-quarter of the sample).

© European Southern Observatory (ESO) 1997

Online publication: June 5, 1998

helpdesk.link@springer.de