Delta Scuti Star Newsletter

Issue 9, September 1995


Cover page, Contents, Previous Page, Next Page

DSSN-Homepage, Author Index, Subject Index


The 1995 FG Vir campaign: will 550 hours of new data and 19 frequencies be enough to do asteroseismology?

by: Michel Breger

For the team: M.Ashley, R.Garrido, G.Handler, L.Mantegazza, A.Nitta, E.Poretti, O.Prouton, E.Serkowitsch, R.Shobbrook, D.Sullivan, W.Zima

Investigators researching Delta Scuti stars have to live with a smaller number of detected pulsation frequencies than the astronomers studying white dwarfs and PG 1159-035 stars. For example, for the star PG 1159-035 over 200 pulsation frequencies are known, while below we are pleased to report a new record for Delta Scuti stars: 19 frequencies for FG Vir. The situation, however, is not as bleak for Delta Scuti stars as it may appear. In white dwarfs, due to the high order of the excited pulsation modes in this star, the values of the frequencies obey the asymptotic limits of frequency spacing. Consequently, a certain amount of redundancy is present. This is not the case for the Delta Scuti stars, where every single newly discovered frequency provides additional, almost independent information on low-order p and g mode pulsation, allowing refinements of the models.

For evolved Delta Scuti stars, the classical picture of separation of the nonradial p and g modes breaks down: in models of evolved stars, the two types of modes are not separated in frequency. Also, the nonradial modes at the frequencies seen in Delta Scuti stars have a mixed character: acoustic (p) in the outer envelope and gravity (g) in the interior. Examples of low-degree p and g modes for a model corresponding to the star FG Vir can be found in a paper by Dziembowski (ASP Conf. Ser. 76,586, 1995).

The Delta Scuti variable FG Vir has been studied extensively during the last few years. During 1992, Mantegazza, Poretti and Bossi (A&A 287, 95, 1994) measured FG Vir photometrically for 8 nights and spectroscopically for one night. They were confident of the correct identification of six frequencies, while a seventh mode of pulsation was also suggested.

During 1993, 170.4 hours of photometric data was obtained for the Delta Scuti variable FG Vir as part of a combined campaign with the Whole Earth Telescope (WET) and the Delta Scuti Network. This allowed the determination of ten pulsation frequencies between 9.19 and 34.12 c/d (112 and 395 micro Hz) with amplitudes between 0.8 to 22 mmag (Breger et al., A&A 297, 473, 1995, hereafter called Paper I). The data suggested the existence of a large number of additional pulsation frequencies. Furthermore, the essentially continuous measurements of WET provided the means to examine the star for the presence of high frequencies in the 1 to 10 mHz range, i.e. in the roAp star range. No variability was detected (A&A, in press).

Without prior knowledge of the pulsation frequencies, the power spectra of 93 hours of photometry obtained earlier during 1985/6 were difficult to interpret without ambiguity. Now, 9 of the 10 frequencies found during the 1993 WET/DSN campaign could also be seen in the 1985/6 data. The comparison indicated that no significant amplitude variability had occurred between the two data sets spaced eight years apart (Dawson, Breger and Lopez de Coca, PASP 107, 517, 1995).

The ten detected pulsation modes could be matched by Dziembowski models computed specifically for FG Vir. A comparison of these models with the observations indicated the presence of both g and p modes. Note that the identification presented in Table 4 of Paper I follows the sign convention used by Dziembowski: n=1 refers to the radial fundamental, rather than the first overtone mode. The comparison between observation and theory also showed that the observed frequencies can be fit by several different pulsation and evolutionary models as well as different values of l for the main variation at 12.72 c/d. A unique identification requires additional frequencies as well as independent determinations of the pulsation quantum numbers l and m.

Consequently, a new multisite campaign of FG Vir was undertaken during 1995 March and April with the Delta Scuti Network. With over 550 hours of photometric observations, this may be the largest campaign undertaken so far and more than all the previous data available for FG Vir. The largest contributions to the campaign were made by Gerald Handler observing at the South African Astronomical Observatory (132 hours on 28 nights) and Michael Ashley from the University of New South Wales (Australia), who measured FG Vir for 141 hours on 30 nights with the APT using a CCD detector. While some astronomers in the field already have attempted, and in some instances achieved, millimag accuracy with CCD detectors, this is the first test with CCD measurements for the Delta Scuti Network. In order to allow an impartial evaluation of these CCD data, for eight nights we have CCD as well as conventional photomultiplier data obtained at another Siding Spring Observatory telescope. The comparison should prove to be very interesting.

At this stage, less than 75% of the photometric data has been reduced. A preliminary power spectrum shows a large number of frequency peaks, of which at least 19 look very promising at this stage. All ten frequencies found in Paper I have been confirmed; this includes the mode at 34.12 c/d. This relatively high-frequency peak had been statistically significant in the combined 1982/3 data set and was therefore included in the solution. However, a subjective impression of its reality had been less positive due to its small amplitude under 1.0 mmag. A number of the additional promising peaks (Paper I, Table 3), but not all, have also been confirmed.

Table 1. Promising peaks in the power spectrum of FG Vir data (preliminary)
cycles per day
9.20
9.66
11.19
11.88
12.15
12.72
16.07
19.23
19.87
20.29
21.05
21.23
21.55
23.40
24.20
24.23
25.44
28.14
34.12

In Table 1 are listing the values of the 19 frequencies shown by the preliminary data. It must be emphasized again that the information is preliminary and may change once the reductions have been completed. Especially the interpretation of the two peaks between 11 and 12 c/d as independent stellar pulsation modes needs to be investigated more. Also, the two close frequencies at 24.20 and 24.23 c/d, respectively, are near the limit of resolution, but are seen in both the 1993 and 1995 data sets.

The determination of the multiple pulsation frequencies is a difficult and challenging undertaking. But to apply asteroseismology, the discovery of the frequencies can only be a first step. The frequencies need to be identified with specific pulsation modes in order to enable a comparison with specific stellar structure and evolution models. Recognition of frequency patterns in power spectra, which works well for stars such as PG 1159-035, is generally not very useful for Delta Scuti stars. The Q values permit the estimate of n, the radial order, but not the degree, l, or azimuthal number, m. This information can be provided by observing phase shifts between different colors, and by examining the behavior of spectroscopic lines. The first method is used by the present photometric campaign, since the stellar brightness variations were measured through the Stromgren v and y filters. The second method will also be applied, because Mantegazza and Poretti were successful in obtaining 26 hours of spectroscopic data on the ESO CAT telescope.

[an error occurred while processing this directive]