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

1. Introduction

The areas in the Hertzsprung-Russell Diagram (HRD) known to contain pulsating stars have continually been increasing both in number and size. Especially the stars with periodic variations have met much observational and theoretical interest because the modelling and identification of the pulsation modes can considerably refine the knowledge of the internal stellar structure (Gautschy & Saio 1995, see also Brown & Gilliland 1994).

In the case of the sun, the seismological approach has been highly successful. However, the demands on the sensitivity to very low-amplitude variations have rendered many attempts to detect oscillations also in other solar-like stars exceedingly difficult. The first significant detection may have been achieved only recently (Kjeldsen et al. 1995).

Towards higher effective temperatures and on the main sequence, the conventional group of pulsating stars closest to the sun is the one of the Scuti stars where the classical instability strip crosses the main sequence. It actually comprises Population I pulsators with radial modes and nonradial low-overtone p -modes (proto-type: Scu), Population II stars with radial modes and larger amplitudes (proto-type: SX Phe), and the magnetic rapidly oscillating Ap stars with nonradial high-overtone p -modes (Matthews 1993). Since in many of these stars several modes are detected, they are suitable candidates for stellar seismology (Däppen 1993) although the density of the eigenmode spectrum makes the mode identification difficult.

Photometry places the cool edge of this strip at 7500 K on the zero-age main sequence, 6950 K at  = 1.7 (Breger 1979). However, not even within this conventional limit all stars are photometrically variable. One possible, partial explanation may be that many Scuti stars are nonradial pulsators, and cancellation effects across the stellar disk would not normally let integral-light photometry detect modes with azimuthal order . (Note that throughout this paper we use the terms 'mode order' and 'order' to be equivalent to the azimuthal nonradial mode index, m, and not to the radial mode order, k or n.) In rapidly rotating, broad-lined stars, the Doppler effect raises this limit up to m  = 12 or more for spectroscopy, depending on v sin i and the local line width. This technique has also been successfully applied to a number of Scuti stars (cf. Yang 1991, Merryfield & Kennelly 1993).

Relatively recently, in this part of the HRD the group of the Dor variables has been defined (cf. Breger 1995; Krisciunas & Handler 1995). They overlap with the cool side of the Scuti variables but candidate members have been claimed down to signifcantly lower temperatures (cf. Breger & Beichbuchner 1996) . More importantly, their photometric periods are up to an order of magnitude longer than those of the radial fundamental mode. In 9 Aur, concomittant variations of radial velocity and spectral line widths and depths support the implied conclusion that the variability is due to nonradial g -modes (Krisciunas et al. 1995); similar evidence has been presented for Dor itself (Balona et al. 1996). By contrast, the evidence for g -modes in Scuti stars is indirect at most (Breger 1993), and in the sun the detection of g -modes is disputed. The only other little to moderately evolved stars with well-established g -mode pulsation are the B- and late O-stars with the 53 Per stars (Smith 1981) and probably the Be stars (Gies 1994; Baade & Balona 1994; but see also Balona 1995). From a seismological point of view, g -modes are interesting because unlike p -modes they probe the stellar interior.

The driving mechanism of the conventional Scuti stars is the classical mechanism working in the zone of partial second ionization of helium. In fact, they are located in the extension of the Cepheid instability strip towards the main sequence. Also for the cooler Dor variables, the He II opacity bump may be responsible; however, convection effects in the driving zone make models numerically delicate, and overstable low-degree modes have not so far been found in them (cf. the remark by Dziembowski in the discussion following Breger 1995; Gautschy & Löffler 1996). Nevertheless, the relation between the Scuti p -mode pulsators and the presumed g -mode Dor pulsators bears some resemblance to the p -mode pulsating Cephei and the g -mode pulsating B stars of somewhat later spectral type (Gautschy & Löffler 1996 and references therein). In both of these two types of more massive stars, the driving is provided by the metal bump in the opacity profile. Whether short-period p -modes or long-period g -modes are excited, depends largely on the thermal time scale of the radial zone where the metal bump occurs (e.g., Moskalik 1995). In the hotter Cephei stars, the relevant temperature zone is closer to the surface so that the thermal time scale is too short for g -modes but compatible with p -mode periods. (Note that this theoretical picture is not fully congruent with the observations as long-period pulsators, the 53 Per stars, have been found on both sides of the Cephei instability strip - see, e.g., Smith 1981.) This scheme might repeat itself for the Scuti p -mode pulsators and the Dor g -mode pulsators where, however, the He II ionization bump would take over the role of the metal bump.

Since stars with sufficiently broad rotational line profiles are found down to late F and even early G spectral types, the Doppler imaging method provides another way of searching for oscillations in stars in the vicinity of the sun in the Hertzsprung-Russell Diagram. Here we report on the results of on attempt of its practical application. Since our observations were planned in 1992, when the notion of the Dor class of variables did not yet exist, we were guided in the choice of our observing strategy by the much shorter periods of high-order modes in Scuti stars and of the solar oscillations. If at all feasible, this required a 4-m class telescope with which, however, not enough time could be expected for the proper sampling of variations with mutliple periods of the order of one day.

© European Southern Observatory (ESO) 1997

Online publication: June 5, 1998

helpdesk.link@springer.de