However, there are still several open questions. For instance, the pulsating DA white dwarfs (DAVs) were for a long time hard to understand, because they do not show many pulsation modes which we can use for asteroseismological anlysis, hence do not provide much information. These objects appear to separate into two groups, the hotter, low-amplitude, rapid (periods around 200 seconds), pulsators and the cooler, high-amplitude, slow (periods around 600 seconds), pulsators. In recent years, considerable insight into the pulsational behaviour of those stars has however been gained by the method of ensemble asteroseismology, i.e. by studying the group properties of pulsating DAVs (Clemens 1993, PhD thesis, University of Texas), or by making use of temporal changes in the pulsation spectra, revealing more and more modes with time (Kleinman et al. 1998, ApJ 495, 424). By performing ensemble studies, one can observe more normal modes to be used for asteroseismological analysis compared to single-star investigations.
In this respect, the situation for the pulsating DB white dwarf stars (DBVs) is not yet clear. Although the study of GD358 by Winget et al. (1994) has provided the most complete picture of a DBV to date, the other representatives of this group are often poorly observed. In addition, mostly high-amplitude DB pulsators have been found, with pulsation periods and amplitudes similar to that of the high-amplitude DAV stars. This situation has changed with the discovery of pulsations in EC 20058-5234 (Koen et al. 1995, MNRAS 277, 913), which turned out to be a the first short-period low-amplitude DB pulsator.
Very recently, a second short-period low-amplitude DBV has been discovered (Handler, in preparation for MNRAS Letters). Just as EC 20058-5234, this star is also rather hot and it seems now that the pulsating DB white dwarfs separate into two groups like the DAVs. This is an excellent starting point to perform an exploration of the group properties of the DBVs. Several questions need to be addressed:
The last question is extremely important for the understanding of white dwarf star evolution in general. It appears that there are two different "channels" through which DB white dwarfs originate, single and binary star evolution. If so, these stars will have different interior structure, a prediction which can be tested through asteroseismology (e.g. see Nitta & Winget 1998, Baltic Astronomy 7, 141).
Regrettably, many DBVs are relatively faint and thus poorly studied. We have therefore started a project to obtain more observational data on these objects, and we have already collected the published and unpublished observations we are aware of. From a reanalysis of these data we concluded that we will need to reobserve four known DBVs (KUV 05134+2605, CBS 114, PG 1456+103 and PG 1654+160), for which mostly only small discovery data sets are available. These four stars can all be measured from SAAO, but we will also involve colleagues at other observatories. For our measurements, the UCT CCD camera is the instrument of choice, as it allows us to observe all our targets with sufficient signal to noise even at moderately-sized telescopes. The remaining DBVs have been studied with the Whole Earth Telescope previously; we will also make use of these data.
With all the data in hand, we will perform uniform period analyses and we will compare the behaviour of the different stars. We will then attempt to arrive at mode identifications for all DBVs, first by applying asteroseismology to the individual stars or attempt ensemble asteroseismology if possible. We are hopeful that this project will let us understand the interior structure (e.g. Helium layer masses) of the DBV white dwarfs as a whole and that of the individual objects and their origin as well.
The case of PG 1654+160:
This proposal asks for including the DBV PG 1654+160 as secondary target in XCOV 21. The star was discovered to be a pulsator by Winget et al. (1984, ApJ 279, L15), who noted its very complex pulsation spectrum with a large range of excited periods. It therefore seemed to be a very promising object for asteroseismological study. The star has however only rarely been observed since. We collected all the available data for the star (one night of the discovery observations has been lost), re-analysed them and show the results in Fig. 1.
The complex pulsational behaviour suggested from the discovery observations is obvious. Even the single light curve obtained in 1984 shows the presence of a multitude of modes. On the other hand, later (unpublished) light curves obtained by S J Kleinman are dominated by a single pulsation mode, although others are still present. In addition, the period analyses of all three available seasons of observing all yield different yearly results, which can also readily be seen from Fig.~1. Finally, PG 1654+160 is, together with the well-studied GD 358 and CBS 114, which will be observed in February 2001, the third proven example of a DBV star with gross temporal changes in its amplitude spectrum. To aid in the understanding of these phenomena, an extensive study of PG 1654+160 is required.
Because of problems with aliasing in the anticipated complicated pulsation spectrum of the star we ask for observations with the Whole Earth Telescope. At this point it is not justifiable to propose primary target status; therefore we suggest to observe it during XCOV 21 after the primary target has set.
Such a strategy was already quite successful during XCOV 20: the secondary target, the DBV KUV 05134+2605, although only observed from practically one geographical longitude turned out to be interesting enough (it is likely to be massive and could therefore have originated from binary star evolution) that a large multisite campaign is currently being organised for next observing season. As more larger telescopes and more Northern sites will be involved in XCOV 21 we hope to be able to obtain a satisfactory understanding of the pulsation spectrum of PG 1654+160 from the low-priority WET observations alone.
Figure caption: Periodograms for all available time-series photometric
data of PG 1654+160. Complex multiperiodic behaviour as well as temporal
changes in the pulsation spectrum are obvious.