In order to study the basic radiative mechanisms in the extended envelopes of evolved stars (e.g. Long Period Variables), we determine the spectrum emerging from a very optically thick Non-LTE hydrogen layer surrounding a core of high temperature (see figure 1). Such a model was first proposed by Sobolev (1960) and Menzel (1946). It is consistent with the fact that LPVs are likely evolving toward planetary nebulae. The main parameters of the model are the temperature of the illuminating star (Tr), the dilution factor (ω) and the Lyman continuum optical depth of the envelope (τ1c). We examine their influence upon the emerging intensities in the lines and in the continua.

The model could explain some observations related to evolved stars and especially the presence of emission features in an otherwise cold environment. Furthermore atmospheric pulsation phenomena produce large variations in the emerging spectra: reddening, emission/absorption transition of the hydrogen lines, profile deformations…

I present first results of a study about propagating radial pulsation modes in the outer atmosphere of Arcturus (K1.5 III). Mechanical energy input is explicitly taken into account by treating shock wave dissipation. I investigate the influence of different wave frequencies on the mass loss behavior of the star. I show that significant time-averaged mass loss can only be produced when periods larger than 5 x 105 s (∼ 1 week) are employed. The initial atmosphere I use extends from 1.2 R* up to 11.8 RR*. All wave models are adiabatic. I found that the mass loss rates and the final flow speeds obtained are extremely sensitive to the wave periods. In a certain regime the effect of increasing the period by a factor of 4 is to increase the corresponding mass loss rate by four orders of magnitude. I note that the mass loss rate and final flow speed of the wind for a 5.6 x 105 s period wave are somewhat close to the observed values. A more complete discussion regarding mass loss generation in Arcturus has been presented by Cuntz (1990). Recent observations of low-amplitude radial velocity variations in the photosphere of Arcturus provide evidence that the theoretically predicted mass loss frequencies might exist. Belmonte et al. (1990) presented evidence for a ∼8.3 d period with an amplitude of ∼50 m s-1, which they attributed to the fundamental radial pulsation mode. Further studies are in progress (Larson et al. 1992). I note that if the 8.3 d period is real than it would be sufficiently large to support continuous mass loss. Judge & Stencel (1991) argued that the mechanical energy in the observed disturbances could be ∼15 times greater than the energy required to drive the wind. For other stars than Arcturus the required minimum mass loss periods can be estimated as PML ∼ R*.

Nonlinear oscillations of proto-giant-planets have been found in recent numerical calculations relevant to planetary formation. Pulsations are excited in two phases of the protoplanetary evolution, (a) In an ‘instability strip’ at core masses of typically 0.2M⊙ (M⊕ is the earth mass). Perturbations grow into the nonlinear domain and saturate into perodic variations with relative luminosity-amplitudes of 0.2m (b) At the so called critical mass (typically at Mcore ≈ 15M⊕). There the pulsations drive a strong mass loss. A large portion of the envelope is ejected. Then the mass loss fades and the envelope settles into a new quasi-equilibrium. This remnant — a post nucleated instability protoplanet — has a compact envelope and is in core and envelope mass similar to Uranus and Neptune.

In this paper we summarize recent work on Cepheid stars by Chiosi et al (1992a), who calculated large grids of models at varying mass, effective temperature, mass-luminosity (ML) relation or equivalently underlying evolutionary scheme for intermediate mass stars, i.e. classical models, models with mild core overshoot, and models with large core overshoot, and finally initial chemical composition. The chemical parameters bracket the abundances of the Cepheid stars in the Milky Way and Magellanic Clouds. First, we present the theoretical results limited to the analytical fits of the fundamental period-mass-radius (PMR) relation and the blue and red edges of the instability strip for the fundamental mode and first overtone. Second, we discuss the data converted to magnitudes and colours of the BVRcIc passbands with particular attention to the period-luminosity-colour (PLC) and period-luminosity (PL) relations. Finally, we briefly summarize three studies based on these models, which aimed to explain the shape and width of the observed instability strip (Chiosi et al 1992b), and the main reason for the mass discrepancy of the Cepheid stars (Chiosi et al 1992c) by means of the mass equivalency (ME) method applied to clusters of the Large Magellanic Cloud (LMC). Because the ME method turned out to be sensitive to the distance modulus, the dependence of this on the metallicity has been the subject of the last study by Bertelli et al(1992).

As tracers of old stellar populations and as primary Population II standard candles, RR Lyrae stars have played an important role in the development of modern astronomy. Our knowledge of stellar evolution has identified these variable stars in a core helium burning phase of low-mass star evolution, the hoiizontal-branch (HB) phase. Consequently, not only to understand fully the nature of RR Lyrae stars, but also to apply them correctly as population probes and distance indicators, we must understand the underlying evolutionary effect of HB stars.

In this paper, I briefly review the most important properties of RR Lyrae stars predicted from the HB evolutionary models, and present many pieces of supporting evidence for these models. For the implications of these models on the chronology of the Galactic formation and on the cosmological distance scale, the reader is referred to several recent publications by Lee (1992a,b,c,d).

Recent results of investigations of the nonequilibrium effects between gas and radiation on the red edge of the Cepheid instability strip are presented.

For the most part it had been assumed that the red and the blue edges of the Cepheid Instability region were parallel. However, previous work by Pel and Lub (1978) and recent work by Fernie (1990) seems to reveal a rather interesting structure to the shape of the Cepheid Instability region. Figure 1 shows the shape defined using the data from Fernie (1990) and the observational data of Gieren (1989). It is apparent that the edges defined by the distribution of these points are not parallel.

I have calculated a series of pulsation models that included varying values of the mixing length to try and produce a distribution of Cepheids as seen in Figure 1. Calculations were done using the methods outlined by Castor (1971) with a Linear Non Adiabatic pulsation code to determine the characteristics for a given model.

The low amplitude radially pulsating classical Cepheid Polaris (α UMi) has well observed decreasing radial velocity and light amplitudes and an increasing period. Recent work indicates that Polaris may stop pulsating by 1995 (Dinshaw et al, 1989), which would make it an excellent indicator for the red edge of the Cepheid instability strip in the Hertzsprung-Russell diagram.

We have calculated Population I envelope models with 200 mass shells of a 5.5 M⊙ star near Teff = 5700 K and log(L/L⊙) = 3.11 which match Polaris’ characteristics (Arellano Ferro, 1984). Models were calculated using the Linear Non Adiabatic pulsation analysis code of Castor (1971). The effects of convection on the pulsational stability were calculated with the time dependent convection theory of Cox, Brownlee and Eilers (1966) as discussed by Cox and Guili (1968). The characteristics of the models were altered until the appropriate period, temperature and luminosity were obtained.

I describe qualitatively the constraints on RR Lyrae masses that arise from consideration of the interior structure of Horizontal Branch (HB) stars. Briefly, the stellar models lying within the temperature range of the instability strip are slightly less massive than those with deep exterior convection zones. I discuss therefore what brings about the existence of these zones in the models.

Horizontal branch (HB) models were evolved using the Yale stellar evolution code, YREC, to test the possibility that mass loss during the RR Lyrae phase is able to produce the observed color (mass) dispersion on the HB (Willson and Bowen 1984) and the anomalous period changes in RR Lyrae stars (Laskarides 1974). Models of total mass 0.64, 0.66, 0.68, 0.70, and 0.72 M⊙ (YMS = 0.23, Z = 0.001) were evolved with constant mass loss rates of 0, 10-10, and 10-9 M⊙ yr-1. Mass loss was assumed to occur only in the RR Lyrae phase, and the instability strip was defined by 3.800 < log Teff < 3.875.

HB stars which lose mass evolve further to the blue. Low mass loss rates do not affect the shape of the tracks significantly. Stars, which without mass loss could not become blue HB stars, were able to emerge from the instability strip on the blue side.

A large number of hydrodynamic pulsation models are converged in the first overtone, and their mean properties compared with observations of RRc stars in six globular clusters. The two observed quantities, period and Fourier parameter ϕ31, lead, via the models, to inferred values for mean mass, luminosity and temperature of the RRc sample in each cluster. We find a narrow range in intracluster RRc luminosity and temperature, but a wider range in mass. At the same time, the intercluster spread is wide in all three parameters. A full discussion of our techniques and results will be given elsewhere (Simon and Clement, in preparation).

The mass of RR Lyrae variables has been a controversial problem for about a decade: while the stellar evolution theory predicts masses ranging between 0.65 and 0.75 M⊙ for Oosterhoff type I and II clusters respectively, the stellar pulsation theory predicts smaller masses (0.55 and 0.65 respectively) using the double-mode pulsators. Simon (1990, M.N.R.A.S. 246, 70), comparing hydrodynamical models with observed stars by means of Fourier parameters, has found relations between the stellar mass and its luminosity, pulsation period, Helium content and Fourier parameter ϕ31. Combining his equations we obtain:

(1)

from which one can estimate the stellar mass by using the observable quantities P and ϕ31 and a luminosity scale, e.g. the one derived by Cacciari, Clementini and Fernley (1992, Astrophys. J. in press).

New CCD BVI observations of RR Lyrae variables in M3 are presented. Mean magnitudes and colours are derived, as well as their relations with periods and amplitudes, and comparisons are made with previous data (Sandage 1981, 1990). Preliminary results are presented on the temperature distribution of the variables and the period-shift effect with respect to M15 and M68.

Using recently published infrared photometry of RR Lyrae stars from both the field (Fernley et al. 1992) and globular clusters (Longmore et al. 1990), period shifts have been calculated using the mean (V – K) colour rather than the mean (B – V) colour employed in most previous analyses (e.g., Sandage 1990 and references therein). The advantage of using (V – K) rather than (B – V) is the reduced sensitivity to, firstly, metallicity and, secondly, non-LTE radiation. This latter occurs in many RR Lyrae stars near maximum light and clearly should not be included when calculating the mean colour and hence, mean temperature. This, and all other aspects of the present paper, are discussed more fully in Fernley (1992a).

The mean (V – K) colour of both the field and cluster stars were de-reddened using the maps of Burstein and Heiles (1982) and converted to effective temperature using the calibration of Fernley (1989).

RR Lyrae stars both in the field and in clusters can be used to derive the metal abundance of the regions and systems where they are found.

(1) New data have been collected on a sample of field ab-type RR Lyraes with the aim of studying the composition of the halo and the disk of the galaxy, (Clementini et al. 1992a, in preparation), using the relation found by Clementini et al. (1991), (hereafter CTM91), between [Fe/H] and the equivalent width of the Ca II K-line W‘(K). (2) A quantitative chemical abundance analysis of the ab type RR Lyrae (V29) in the globular cluster M4 has been performed using high resolution, high S/N spectroscopy. We obtain [Fe/H]=–1.3 ± 0.2 and the α– elements (Mg and Ti) are overabundant by 0.6 dex. These results are in good agreement with determinations from high resolution spectra of giants and blue horizontal branch stars (Clementini et al. 1992b, in preparation).

A search for variable stars in the globular cluster NGC 6544 has revealed only one possible short period variable within the tidal radius of the cluster. A search in NGC 6642 yielded 16 new RR Lyrae stars within the tidal radius and 5 new field RRs. The previously discovered (Hoffleit 1972) V1 is a slow variable, and V2 is an RR Lyrae star. Photometry of the variables within the tidal radius gives a mean B for the horizontal branch of < B > = 17.0 mag. With E(B – V) = 0.37 mag and (B – V) = 0.35 mag for RR Lyraes, a value for V(HB) = 16.3 mag is derived. This is about one mag fainter than previous estimates (Webbink 1985), and places NGC 6642 at a distance of approximately 7.9 kpc.

We present preliminary results of a program of CCD photometry of RR Lyrae variable stars in the globular clusters NGC 6388 and M15.

Stellingwerf's one–zone model is a simple and useful tool for reproducing the main observed pulsational characteristics of RR Lyrae and high amplitude δScuti stars, in particular their light and color curves (Stellingwerf et al., 1987, Ap.J.313, L75; Antonello, 1990, Astr. Ap.230, 127). In the present poster we show in better detail a comparison of the observed amplitudes of the light curve at various wavelengths with those predicted by the one–zone model; a preliminary result on this subject was reported in the short note by Grieco and Antonello (1990, in Confrontation between Stellar Pulsation and Evolution p. 101). As in the previous applications, here we use the published grids of atmospheric models (Kurucz, 1979, Ap.J.Suppl. 40, 1) and we do not consider possible shock effects.

The AF stars are those of spectral types A and F that have been discovered on objective-prism plates taken with the Burrell Schmidt in the Case Low-Dispersion Northern Sky Survey (Pesch and Sanduleak, 1983). In SA 57, this survey is complete to V = 16.5. In this field and also in the Lick Astrograph RR Lyrae survey field RR 7 (in the anticenter) the AF stars comprise (a) blue horizontal branch stars and RR Lyrae stars of the halo and (b) stars which have the higher surface gravities of main sequence stars. The two groups can be separated primarily by their differing Balmer jumps and Balmer line-widths. The latter group (which may well include blue stragglers of both Pop I and Pop II) extends some 10 kpc above the galactic plane and shows a wide range of metallicity. Photoelectric photometry of this AF star sample has allowed the detection field RR Lyrae stars of lower-amplitudes than could have been found by conventional blinking techniques; this has led to a significant increase in the number of RR Lyrae stars that are known in SA 57 and RR 7. The cooler main sequence and/or blue straggler AF stars lie in the zone of pulsational instability and one higher-amplitude δ-Scuti star was detected in field RR 7. It is suggested that these AF stars provide a good sample for studying the incidence of pulsation in the population of older stars that extends beyond the thin disk.

We report results of a new photographic survey of variable stars in a 1 x 1.3 degree region near the Northeast Arm of the Small Magellanic Cloud. We have discovered 133 new variable stars in this field and have determined periods and B lightcurves for 78 new and 72 previously known variables. At periods shorter than about 3 days, the Cepheid period-luminosity relation splits into two sequences. The brighter sequence is believed to be populated by stars pulsating in the first overtone radial mode, wherecis the fainter sequence is populated by fundamental mode pulsators. The peak in the Cepheid period-frequency distribution occurs near a period of 1.8 days. The surface density of RR Lyrae stars in this field is comparable to that in an outlying SMC field near NGC 121.