MSST observations of the pulsating sdB star PG1605+072

Abstract

We present the first results from the MultiSite Spectroscopic Telescope (MSST) observations of the sdBV star PG1605+072. Pulsating sdB stars (also known as EC14026 stars) offer the chance to gain new insights into the formation and evolution of extreme Horizontal Branch stars using the tools of asteroseismology. PG1605+072 is an outstanding object in its class, with the richest frequency spectrum, the longest periods, and the largest variations. The MSST campaign took place in May/June 2002 immediately following the Whole Earth Telescope Xcov22 run, which observed PG1605+072 as an alternate target. We will first give an overview of the project and its feasibility, after which we will present the massive data set, made up of 399 hours of photometry and 151 hours of spectroscopy. The overall aims of the project are to examine light/velocity amplitude ratios and phase differences, changes in equivalent width/line index, and wavelength-dependence of photometric amplitudes, and to use these properties for mode identification. Comment: 8 pages, 4 figures, includes Kluwer class file, to appear in "Extreme Horizontal Branch Stars and Related Objects", proceedings of the meeting held in Keele, UK, June 16-20, 2003

Full text

Page 1

arXiv:astro-ph/0309061v1 2 Sep 2003
MSST observationsof the pulsating sdB star PG1605+072
S. J. O’Toole (otoole@sternwarte.uni-erlangen.de), S. Falter and
U. Heber
Dr Remeis-Sternwarte, Astronomisches Institut der Universit¨ at
Erlangen-N¨ urnberg, Sternwartstr. 7, Bamberg D-96049, Germany
C. S. Jeffery
Armagh Observatory, College Hill, Armagh BT61 9DG, Northern Ireland, UK
S. Dreizler and S. L. Schuh
Insitut f¨ ur Astronomie und Astrophysik, Universit¨ at T¨ ubingen, Sand 1, D-72076
T¨ ubingen, Germany
Universit¨ atssternwarte, Universit¨ at G¨ ottingen, Geismarlandstrasse 11, D-37083
G¨ ottingen, Germany
and the MSST+WET Teams∗
Abstract. We present the first results from the MultiSite Spectroscopic Telescope
(MSST) observations of the sdBV star PG1605+072. Pulsating sdB stars (also
known as EC14026 stars) offer the chance to gain new insights into the formation
and evolution of extreme Horizontal Branch stars using the tools of asteroseismol-
ogy. PG1605+072 is an outstanding object in its class, with the richest frequency
spectrum, the longest periods, and the largest variations.
The MSST campaign took place in May/June 2002 immediately following the
Whole Earth Telescope Xcov22 run, which observed PG1605+072 as an alternate
target. We will first give an overview of the project and its feasibility, after which
we will present the massive data set, made up of 399 hours of photometry and 151
hours of spectroscopy. The overall aims of the project are to examine light/velocity
amplitude ratios and phase differences, changes in equivalent width/line index, and
λ-dependence of photometric amplitudes, and to use these properties for mode
identification.
Keywords: stars: oscillations — subdwarfs — stars: individual: PG1605+072
1. Motivation
Of all the short-period sdB pulsators currently known, PG1605+072
stands out. It is perhaps the most evolved, it has the highest amplitudes
and longest periods, and it is the only apparently single sdB star that
shows significant rotation. Its rich pulsation spectrum and relative
brightness make it an ideal target for multisite observing campaigns
such as the Whole Earth Telescope (WET). After the discovery obser-
vations showed a complex amplitude spectrum with many modes (Koen
∗see http://astro.uni-tuebingen.de/∼schuh/msst/astronomers.html and
http://wet.iitap.iastate.edu/xcov22/people.html
c ? 2008 Kluwer Academic Publishers. Printed in the Netherlands.
KeeleTalkProc.tex; 2/02/2008; 6:11; p.1

Page 2

2
S. J. O’Toole et al.
et al. 1998), a two week multisite photometric campaign was organised,
and more than 55 frequencies were detected, as well as evidence for
amplitude variation (Kilkenny et al. 1999). Models by Kawaler (1999)
suggested that part of the complexity of PG1605+072’s amplitude
spectrum may be due to rapid rotation with an equatorial velocity of
about 130kms−1, thereby causing rotational splitting of the oscillation
modes. When Heber et al. (1999) carried out a spectral analysis, they
measured vsini to be 39kms−1, which fit nicely into this picture. Both
Heber et al. (1999) and Koen et al. (1998) found that PG1605+072
has evolved off the extreme Horizontal Branch (EHB), making it an
important link between the EHB and the white dwarf cooling curve.
An analysis by Reed (2001) showed that the amplitude spectrum is too
complex to measure mode stability.
O’Toole et al. (2000) were the first to detect velocity variations in
the Balmer lines of PG1605+072, measuring amplitudes of 14kms−1
in Hβ using 2m telescopes. Woolf et al. (2002) showed the advantage of
using 4m class telescopes, with much better velocity accuracy than 2m
telescopes. They also used moments of the cross-correlation function
to detect line shape variations. O’Toole et al. (2002), using a larger
data set than O’Toole et al. (2000), did a detailed analysis of two-site
spectroscopy, and found evidence for amplitude variation and closely
spaced modes. O’Toole et al. (2003) examined Balmer line indices and
found an amplitude dependence on Balmer line number, which they
used to derive the amplitudes of the effective temperature and surface
gravity variations. Falter et al. (2003) were the first to attempt simul-
taneous multicolour photometry and spectroscopy, and found no phase
difference between different filters. Using spectrophotometry, O’Toole
(2003) obtained the same result, and found the velocity/intensity phase
difference to be ∼110◦for all 8 modes detected (for purely adiabatic
oscillations, this phase difference should be 90◦).
These analyses showed that one and two site spectroscopic cam-
paigns are not enough to understand the complex nature of PG1605+072,
and that even multisite photometry would not do the job. Considering
the potentially huge amount of information that can be obtained from
this star, we decided to undertake a multisite spectroscopic and photo-
metric campaign in May/June 2002. Here we present the initial results
from this ambitious project.
2. Observations and Reductions
The MSST campaign obtained both photometry and spectroscopy. The
photometric part of the campaign was divided into two parts. As part
KeeleTalkProc.tex; 2/02/2008; 6:11; p.2

Page 3

The pulsating sdB star PG1605+072
3
Figure 1. Velocity curve from the Calar Alto 3.5m with the TWIN spectrograph.
Variations with a period of ∼8 minutes are clear.
of the WET Xcov22 campaign, PG1605+072 was observed as an al-
ternative target, and ∼127 hours of observations were obtained (Heber
et al. 2003). All photomultiplier (PMT) data were reduced using the
WET reduction software QED. Most CCD data were reduced using
standard aperture photometry routines in IRAF. Some data remains
unreduced. From the main part of the MSST campaign ∼272 hours of
observations were obtained, giving a total of ∼399 hours, or roughly
54% temporal coverage. This is by far the most data acquired for any
sdB during a single observing campaign. Again PMT data were reduced
using QED, while most CCD data was reduced using TRIPP (Time Re-
solved Imaging Photometry Package, see Schuh et al. 1999), and some
data were reduced using IRAF. We also have 5 nights of multicolour
photometry, using BUSCA on the Calar Alto 2.2m telescope, which
has yet to be reduced.
There were also two parts to the spectroscopic contribution to the
MSST campaign, using 4m and 2m telescopes. Several of the 4m tele-
scopes applied for did not receive time or were clouded out, so coverage
was poor at only ∼7% (around 27 hours total). All 3 telescopes (Apache
Point 3.5m, Calar Alto 3.5m and the ESO NTT 3.5m) acquired spectra
by trailing the star along the slit. This allows for variable exposure
times, depending on conditions. An example of the quality of data
possible using this technique is shown in Figure 1. The high velocity
precision achievable with this technique suggests that it can be used
KeeleTalkProc.tex; 2/02/2008; 6:11; p.3

Page 4

4
S. J. O’Toole et al.
Figure 2. Coverage of the 2m spectroscopy part of the campaign. The 2 weeks gap
in the observations causes ∼0.8µHz alias peaks as well as those caused by daily
gaps.
on fainter and/or lower amplitude targets. Not all of the data has been
analysed, although everything has been reduced using SPEX (long-
slit SPectrum EXtraction package1), a package which allows for the
reduction of trailed spectra. Poor weather conditions during the NTT
observations means that data may not be useful.
The observations using 2m telescopes were somewhat more success-
ful, although the timing of the allocations was not always optimum,
meaning there were two halves to the campaign, separated by about 2
weeks. The first half of the campaign had good coverage on paper,
but bad weather at both La Silla and Siding Spring Observatories
meant around 70-75% of allocated time was lost, leading to 58 hours of
observations or ∼22% coverage. The second half of the campaign was
much more successful, with 93 hours of observations, or around 32%
coverage. All of the 2m spectroscopy data were reduced using standard
routines in IRAF for bias subtraction, flat fielding, sky correction, and
order extraction, however the velocities were determined using a double
precision version of the rv package2. The raw velocities are shown in
Figure 2. The higher apparent scatter in the second half of the campaign
is mainly due to long term drifts in the velocity curves. These drifts
seem to be inherent in both the DFOSC and ALFOSC spectrographs. A
1see http://astro.uni-tuebingen.de/∼schuh/spex/index.html
2available from http://iraf.noao.edu/scripts/extern/rvx.pl
KeeleTalkProc.tex; 2/02/2008; 6:11; p.4

Page 5

The pulsating sdB star PG1605+072
5
Figure 3. Amplitude spectrum of MSST photometry from SAAO, SSO and the JKT.
The 2075.8 µHz mode is dominant again. The inset shows the spectral window.
similar problem is seen in observations of PG1325+101 using ALFOSC
(Østensen, these proceedings).
3. First Results
Although great care was taken to make sure that the timing of each
observation was accurate, inevitably we ran into some problems. These
mainly occurred during analysis of the photometry, where the addition
of several sites to the main campaign data caused strange aliasing
effects. In some cases a large reduction in oscillation amplitudes was
also seen (up to ∼25%), despite the data from single sites analysed
individually showing similar amplitudes. This might indicate a timing
problem. As a start, we show in Figure 3 the amplitude spectrum of
3 sites where timing does not seem to cause problems (SAAO, JKT
and SSO). Fortunately PG1605+072 was observed for at least 6 nights
from each of these observatories, constituting a large fraction of the
data. Other sites appear to have deviating timing, and the reasons are
still under investigation. As mentioned above, some data (from BAO)
have not been reduced yet, and are not included.
One of the most striking things about Figure 3 is the dominance
of the mode at 2075.8µHz. This mode had the highest amplitude in
the observations of Koen et al. (1998) and Kilkenny et al. (1999), how-
KeeleTalkProc.tex; 2/02/2008; 6:11; p.5