During the past several years, asteroseismologists of the Instituut
voor Sterrenkunde of the K.U.Leuven, of the Institut d'Astrophysique et
Géophysique of the ULg and of the Royal Observatory have very
succesfully collaborated on many different topics in asteroseismology
(we refer to the publication list in the annual
reports of the previous IUAP). The collaboration is now extended to the
Observational Astronomy Group of the V.U.B. This implies the participation of
all the members of the Belgian Asteroseismology Group (BAG). The BAG was founded
in 2000 with the specific aim to coordinate all the Belgian initiatives and
expertises in asteroseismology, in order to orient the relevant research towards
the upcoming data of pulsating stars from space. It mostly concerns COROT, a
French-led European mission, which is in full preparation at present, with
launch foreseen in 2005 and lifetime of about two years. Several
members of the BAG are involved in the development of these missions.
1) Hare and Hound exercises for the COROT mission
2) A Cephei stars
In the context of the preparation of the COROT space mission, "hare-hound" exercises were independently performed by several teams of the COROT Seismology Working Group. The aim of such exercises, which are commonly used in helioseismology, is to prepare the data analysis and scientific interpretation of the asteroseismic COROT observations. This will help in the development and testing of the tools which are needed for the interpretation.
The exercise was divided in four consecutive steps, each performed independently by a different group:
For the COROT Week 3, which was organized in Liège, we joined the "hare and hound" exercice at step 4, from a different angle than the other teams. We compared our numerical stellar evolution code and our oscillation code to those used by other (French) teams. By comparing the results of different "theoretical" codes, it is possible to get an estimate of the "theoretical uncertainty" on the results. This is of course essential for the interpretation of the future seismic data.
We found that for the solar-type star chosen for the exercise, we had an almost perfect agreement with the results of the French Team, at least for the evolutionary track of that star in the HR-diagram, and for the oscillations of the star. We had to use slightly different parameters for the star. This will need more in-depth analysis.
We are currently runninghave run two BAG COROT H&H exercises for B stars, the results of which have been presented at the 4th COROT held in June 2003 in Marseille. The conclusion of the COROT team was that a Beta Cephei star should indeed be included in the primary targets of the COROT mission.
3) B and SPB stars
A seismic modelling of the massive star 16 Lac has been made. We have compared a new spectroscopic mode identification with a photometric identification based upon a non-adiabatic description of the eigenfunctions in the star atmosphere. Numerous theoretical models have been computed in order to select the best model fitting the observed frequency values as well as the mode identifications. The derived mass is 9.62 M with an age of 15.7 million years.
We have obtained photometric data assembled by the NASA space mission WIRE of the Cephei star beta Crucis (Mimosa). A total of 5 million observations covering 17 days was analysed and the three main periods found in this way are in perfect agreement with the results derived earlier from line profile variations. The photometric amplitudes are small (3, 2.7 and 0.6 millimag for the dominant modes), but this is not surprising in view of the mode identifications derived earlier from the line profiles. Additional periods of low-amplitude modes (between 0.2-0.3 mmag) are also derived, including one suggested earlier by the radial velocity data.
We have performed an observational study of the orbital motion and the intrinsic variability of the double-lined spectroscopic binary Cen. Using 463 high signal-to-noise, high-resolution spectra obtained over a timespan of 12 years it is shown that the radial velocity of Cen varies with an orbital period of 357.0 days. We derive for the first time the orbital parameters of the system and find a very eccentric orbit (e=0.81) and similar component masses with a mass ratio M1/M2=1.02. Cen forms a challenge for current evolution scenarios in close binaries and it is also a puzzle how a massive binary with such a large eccentricity could have formed in the first place. Both the primary and the secondary exhibit line-profile variations. A period analysis performed on the radial velocity variations of the primary after prewhitening the orbital motion leads to the detection of at least 3 pulsation frequencies while the star does not show any periodic photometric variability.
We try to model Eri by fitting the four most significant, independent frequencies that are present in both the photometric and spectroscopic data. So far we encounter a puzzle : a non adiabatic analysis of the solutions best fitting the observed frequencies shows that these modes are not excited. We now study the influence of a different mixture of heavy elements.
HD 92024 is an eclipsing binary with a b Cephei primary. By combining the orbital and pulsational information from this system, a strong seismic analysis is possible if the multiple modes of oscillations can be identified. The bulk of observations have been collected during the last 15 years, by collaborators from three IAP network partners, at ESO, La Silla with in particular 4-band photometry from the Strömgren Automatic Telescope (SAT). The most recent spectroscopic observations were collected with the FEROS spectrograph during 2001-2002, which brings the total number of spectra above 130, while the last lightcurve data from the SAT rounded 1200 in the year 2003. The time-series spectroscopy shows profound LPVs in the single-lined spectrum due to the pulsations in the Cephei and we find indications of three pulsation frequencies in the radial-velocity data. Because radial velocities (RVs) from spectral lines were strongly affected by line-profile variations (LPVs), a novel method was developed to measure the orbital RVs. By matching pairs of spectra with similar LPV patterns and cross-correlating them, it was possible to get RV differences across the full orbital phase. With a software routine designed by our Croatian collaborators, a Keplerian eccentric orbit was then iteratively optimised to fit the RV differences. This approach significantly reduced the influence of the LPVs on the orbital solution and resulted in an improved orbital period, a small, but significant, eccentricity (e=0.03), and stellar masses of 15 M and 3 M with corresponding radii of 8.3 R and 2.1 R , and temperatures of 25500 and 12500 K. In the course of 2004, the orbital and dimensional analysis will be concluded. Then the orbital imprint will be removed from the spectra and the light curves (with more out-of-eclipse observations added). We will then attempt to detect the secondary spectrum and to identify the pulsation modes. Finally, if successful in the latter, an asteroseismological analysis will be performed in order to gain insight into the stellar structure and other physical properties of the pulsating star.
A full dynamical analysis of the binary was performed. We find stellar dimensions of 15 and 3 solar masses, radii of 8.4 and 2.1 solar radii and temperatures of 25500 and 12500 K for the bright B1III and faint B7-B9V components. The major limitation is the absence of a secondary spectrum. A search for line-profile variations in all lines in the wavelength region from H-alpha to H-epsilon indicated that about 30 lines contain significant pulsational information. It was demonstrated that by combining power spectra across more lines, the detection of oscillation frequencies was enhanced significantly. A ROB-KUL-VUB collaboration performs now a pulsational analysis of the 103 FEROS spectra with the purpose of identifying modes of two or all three known oscillation frequencies. The high s/n averaged spectra were released in a data paper.
We have gathered and analysed a timeseries of 1493 high-quality multicilour Geneva photometric data of the B3V Cep star HD 129929. The dataset has a time base of 21.2 years. The occurrence of a beating phenomenon is evident from the data. We find evidence for the presence of at least six frequencies, among which we see components of two frequency multiplets with an average spacing of ~0.0121 cd-1 which points towards very slow rotation. This result is in agreement with new spectroscopic data of the star and also with previously taken UV spectra. We provide the amplitudes of the six frequencies in all seven photometric filters. The metal content of the star is Z = 0.018 ± 0.004. We then performed a detailed seismic study of the Cep star HD 129929. Our analysis is based on the recent derivation of six pulsation frequencies. These frequencies are unambiguously identified from the seismic modelling and the photometric amplitudes to be the radial fundamental, the l = 1, p1 triplet, two consecutive components of the l = 2, g1 quintuplet. A non-adiabatic analysis allows us to constrain the metallicity of the star to Z [0.016,0.022]. In addition, the fitting of three independent frequencies, two of which correspond to axisymmetric (m = 0) modes, allows us to constrain the core overshooting parameter to OV = 0.10 ± 0.05, as well as the other global parameters of the star. Finally, from the observation of the l = 1 triplet and part of the l = 2 quintuplet, we derive constraints on the internal rotation of this star.
Kappa Sco is a binary Beta Cephei that is known to be multiperiodic. In two short data sets (separated by 152 days) obtained by the star camera of the satellite WIRE the known main period of 4.8 hours is clearly present. The secondary period could also be found, but more periodicities seem to be present. A new analysis of high resolution spectral data lead to a better solution of the parameters of the binary and the first attempts to mode identification were made.
4) Scuti and
In the framework of a long-term spectroscopic and photometric monitoring of slowly pulsating B stars we studied thoroughly the northern target star HD 147394. We performed an end-to-end analysis, consisting of a frequency analysis, a mode identification from line-profile variations and a comparison with theoretical pulsation models.
We present ground-based multi-colour Geneva photometry and high-resolution spectra of four variable B-type stars: HD 105382, HD 131120, HD 138769 and HD 55522. All sets of data reveal monoperiodic stars. A comparison of moment variations of two spectral lines, one silicon line and one helium line, allows us to exclude the pulsation model as being the cause of the observed variability of the four stars. We therefore delete the four stars from the list of candidate slowly pulsating B stars. We attribute the line-profile variations to non-homogeneous distributions of elements on the stellar surface and we derive abundance maps for both elements on the stellar surface by means of the Doppler Imaging technique. We confirm HD 131120 to be a He-weak star and we classify HD 105382, HD 138769 as new He-weak stars. HD 55522 has the solar helium abundance but the mean abundance value of He varies by 0.8 dex during the stellar rotation. For HD 131120 and HD 105382, helium is enhanced in regions of the stellar surface where silicon is depleted and depleted in regions where silicon is enhanced.
We want to investigate: (1) to which extend current disentangling codes can be used for systems containing a pulsating component, (2) if the results of the pulsational analysis can be improved by first subtracting the contribution of the non-pulsating components from the spectra. Our investigation is based on observed high-resolution spectra of HD 140873 and HD 123515 obtained with the CAT/CES combination at La Silla in 1996-1998, and on synthetic data. HD140873 and HD123515 are both non-eclipsing double-lined binaries in eccentric orbits of 39 and 26 days. The broad-lined primary of HD140873 is an apparently monoperiodic SPB, while the sharp-lined primary of HD123515 is a multiperiodic SPB prototype. The spectral disentangling seems to be working fine in the case of HD140873. After removal of the secondary from the data, we re-analyzed the pulsations of the primary, which leads to a better agreement in the temporal behavior of the two SiII profiles. For HD123515, spurious features appear in the disentangled component spectrum for the secondary. We suspect that these features are induced by the pulsations of the primary, which is supported by the tests we did so far with similar synthetic data sets. However, more tests on synthetic data are needed to understand under which conditions the disentangling of the component spectra is applicable to binaries with a SPB component. Presently, two IAP partners are involved in this study.
5) Solar type stars
Spectroscopic and multicolour photometric evidence was presented, supporting the binary nature of the Scuti star XX Pyx. Applying a cross-correlation technique to the spectra, we found clear radial-velocity variations with a large amplitude. We derive the orbital parameters and confirm an orbital period of 1.15d, as suggested previously on the basis of photometric variations. The amplitude of the slow variations present in our new multicolour data is wavelength independent, pointing also to a geometric effect as origin of the variability. They are thus fully consistent with the spectral variations and are interpreted as ellipsoidal variations. XX Pyx has a circular orbit of which the radial-velocity variations have a semi-amplitude of 17.8+/-0.4km s-1. The single-lined binary nature of the star, together with the mass function, lead us to conclude that the orbital inclination must be larger than 10 degrees. The orbital solution is compatible with a synchronized M3V companion. The deformation of the primary due to tidal forces is very probably the reason for the failure of detailed seismic modelling efforts done recently.
New theoretical instability strips for Sct and Dor stars are presented. These results have been obtained taking into account the perturbation of the convective flux following the treatment of Gabriel (1996). For the first time, the red edge of the Sct instability strip for non-radial modes is obtained. The influence of this time-dependent convection (TDC) on the driving of the Dor gravity modes is investigated. The results obtained for different values of the mixing-lenght parameter are compared for the Dor models. A good agreement with observations is found for models with between 1.8 and 2.0.
We have reported on our results from a large photometric campaign on thirty five Dor star candidates undertaken in the framework of the Flanders - South-Africa project. An overview of the data, as well as the results of the analysis of the obtained time series are presented, the main conclusion being that nine stars are thought to be multiperiodic gamma Dor stars and eight monoperiodic. We also performed a photometric mode identification for two stars of the sample by comparing the amplitude ratios in the different passbands of the Geneva photometric system. Both stars seem to pulsate in non-radial modes of degree l = 1.
DG Leo is a hierarchical triple with all three components situated in the Scuti instability strip. Previously, we showed that it consists of a close binary with two mild Am stars, and a Scuti type wide companion showing multi-mode line-profile variability. The system puts empirical constraints on the development of pulsations in stars of the same mass and age, but with different internal structure that provokes different chemical composition in the outer layers. In 2004, a detailed abundance analysis confirming the enhanced metallicity in the close binary and solar abundances in the wider component was performed and published.
6) Variable stars in clusters
Models of Cen A & B have been computed using a new mass determination together with seismological data obtained very recently. These new data do help improve our knowledge of the evolutionary status of the system. All the constraints are satisfied with a model which gives an age of about 6 Gyr for the binary.
Variable stars in star clusters share common ages, original chemical composition and distance, which strongly simplifies the comparison with theoretical models. We take advantage of this when studying pulsating variables in the globular cluster Centauri (though metal abundance is a free parameter in this case), a collaboration between two of the IAP partners and colleagues in Rome (OAR), Germany (AIP) and Denmark (CUAO), and in a second project where we search for open clusters well suited for multi-site campaigns of Sct stars.
7) A and F stars
A photometric investigation is carried out on Centauri with the final purpose of detecting and studying variables (in particular SX Phe stars suitable for asteroseismology), and to study the complex evolution history of this cluster. The observational basis is FORS observations (Chile, Paranal) obtained at high angular resolution (seeing better than 0.3 arcseconds), about 5000 times series points from the Danish 1.5 meter (optical, La Silla), the 3.5 meter NTT (near-infrared, La Silla) and observations from the ESO archive (optical/near-infrared, WFI/2.2m, HST, VLT).
Due to Cen's uniquely large spread in chemical abundances, 3-4 separate red giant branches have been discovered in the last 6 years. One of these branches deviates strongly from the bulk of the cluster stars and the origin for the stars in this branch is particularly puzzling as they appear to be chemically and kinematically different from the rest of the cluster stars. Based on optical (VLT and HST) and NTT (near-infrared) data, we show, by comparing observations with stellar evolutionary models, that this branch and a newly discovered sub-giant branch can be described altogether as a clump of the cluster's stars positioned 500 pc beyond Cen. If confirmed, this supports other indications of presence of a cluster tidal tail, typical for captured dwarf galaxies.
Our separate study of deep hst-Acs photometry resu1ted in the detection of the largest known population ever of white dwarfs (WDs) exceeding 2000 such objects. We find this sample in agreement with predictions based on the ratio between WDs and Horizontal Branch evolutionary lifetimes.
With the Nordic Optical Telescope (NOT, Canary Islands), selected northern open star clusters are examined for their content of short-period variables of the types Scuti and Cephei, currently the most successful classes of variables applied in asteroseismic tests. Photometry of clusters with several of such variables inside convenient field-of-views is then followed up by spectroscopic studies, and photometric multi-site campaigns, which is particularly feasible on the northern hemisphere. The current cluster monitored with the NOT, NGC 1817, was a cluster already known to host 7 potential Scuti stars, and in collaboration between the two IAP partners in Brussels and Aarhus University (Denmark, IfA), 5 nights in December 2002, were used to collect time-series observations in B and V as well as uvby standard data. Preliminary results from the analysis of these observations, presented in Arentoft et al (2004a , 2004b), announced the detection of 14 new variable stars in the cluster. This brings the total number of known variable stars in NGC 1817 up to 19, including 12 multi-periodic Scuti stars. In one case, a Scuti star is also member of an eclipsing binary system, which is valuable for obtaining precise, independent stellar parameters. Calibration of the broadband photometry is now almost finished and is used to confirm the location of the new variables inside the Scuti instability strip, while the standardisation of the uvby data is still in progress. The open cluster studies run in three stages on more clusters: 1) identification of clusters hosting several short-period variables, 2) through uvby photometry and spectroscopy derive physical stellar parameters required for selection of variables suitable for seismic analysis, and 3) a multi-site campaign on the cluster variables in order to reduce the power of alias frequencies in frequency analysis.
8) Non-adiabatic analysis
The Mercator telescope is a new 1.2 meter telescope located on the Roque de los Muchachos observatory on La Palma, Spain. The scientific observations started in spring 2001. Currently, the instrument attached to the telescope is the "P7" which is a two channel (star + sky) photometer for quasi-simultaneous 7 band measurements in the Geneva photometric system. Since 2001, the telescope has been intensively used to observe pulsating B, A, and F main sequence stars. We have at present 31 data sets of A and F stars, containing mainly Doradus stars and 28 data sets of B stars. All data sets were subjected to an extensive frequency analysis, where objectivity was imposed by having each star independently analysed by two or three researchers. The frequency analysis was done with the PDM method and with the Lomb-Scargle method. We find that most stars are clearly multiperiodic. From these results, the most promising targets have been selected for long-term monitoring.
The study of the non-adiabatic character of stellar oscillations can give precise informations about their internal structure. First, we studied in details the influence non-adiabatic temperature variations on line profile variations of B-type pulsators. Our non-adiabatic pulsation code also allows us to determine the theoretical amplitude ratios and phase differences of the light curve as seen in different colour filters. By confronting these theoretical predictions to the observations, it is possible on one hand to determine the degree l of the modes and, on the other hand, to obtain precise informations about the stars. We call this method Non-adiabatic Asteroseismology. We applied this method to the study of B-type pulsators and showed that strong constraints on the metallicity of these stars can be obtained by this approach. We applied also this method to the study of A- and F-type pulsators, in a close collaboration with the Instituto de Astrofisica de Andalucia, Spain. We showed that the characteristics of the thin convective envelope of these stars can be constrained with this approach. Our last works in this frame have been to include the new ATLAS9 atmosphere models in our non-adiabatic code. These models have been computed using three different convection treatments (mixing length theory -MLT-, Canuto and Mazzitelli (1991), and Canuto, Goldmann and Mazziteli 1996) and therefore allow us to compute complete (atmosphere and interior) stellar models with a consistent treatment of convection.
9) Spectroscopic analysis
We implemented a new version of the moment method to identify modes of pulsating stars from high-resolution spectroscopy. This new version is generalised to rotating stars and is more appropriate to multiperiodic stars than any of the other methods available thus far. We have applied the new method to several pulsating B stars and will continue to do so in the near future.
The interpretation of the asteroseismological data requires a precise knowledge of the basic stellar parameters: mass, effective temperature, luminosity, chemical composition. We are coordinating the spectroscopic analyses of the main COROT targets, in order to determine these parameters as accurately as possible. We have also started a detailed analysis of our own, on the basis of high resolution and high signal-to-noise spectra obtained at several observatories by the COROT team.
By testing and comparing different spectroscopic analysis methods, we have been able to implement a method which gives results of unprecedented accuracy. The effective temperature is determined within +/- 30K for the coolest program stars (i.e., the ones not much hotter than the sun) and the accuracy of the metallicity is now mostly limited by the uncertainties in the chemical composition of the sun.
In order to obtain accurate atmospheric parameters and chemical composition for the rapid rotators, we have devised a new spectroscopic analysis technique, based on synthetic spectra. This new method works iteratively. It includes the determination of the continuum level on the basis of pseudo-continuum windows. Basically all spectral lines are blended because of the rotational broadening. We have thus developed a method based on the fitting of these line blends, and using blends with different sensitivities to the atmospheric parameters to determine the effective temperature, surface gravity and chemical conposition. It allows to obtain results of good precision, even for stars with rotation velocities larger than 50 km/s.
The European Network of Excellence in AsteroSeismology (ENEAS) was created on 11 October 2002 during a kick-off meeting held in Leuven (Belgium) at the initiative of the Belgian Asteroseismology Group. The purpose of ENEAS is to provide an efficient coordination between the numerous, already existing European initiatives in asteroseismology and to achieve an effective integration of all European expertise in this research domain through numerous exchanges of the ENEAS scientists. Another prime goal of ENEAS is to prepare European scientists for the huge databases of seismic data of stars that will be delivered by European space missions, such as COROT and Eddington, in the near future. Any information on ENEAS can be found at the ENEAS web page: http://www.eneas.info
This web page will gradually include more and more applications, such as basic information about ENEAS, databases of seismic observations, a database of analysis tools with user-guidelines, outreach material, asteroseismology courses etc. Besides this electronic communication channel, we will publish any ENEAS matters in the journal Communications in Asteroseismology under editorship of M. Breger (Vienna University).
Meanwhile, 256 scientists from 43 institutes have joined ENEAS.