arXiv:astro-ph/0308294 v1 18 Aug 2003AGE DEPENDENCE OF THE VEGA PHENOMENON:OBSERVATIONSG. DecinInstituut voor Sterrenkunde, KULeuven, Celes tijn enlaan 200B, B-3001 Le uven, [email protected]. DominikSterrenkundig Instituut ‘Anton Pannekoek’, Kruislaan 4 03, NL-1098 SJ Amsterdam, [email protected]. WatersInsituut voor Sterrenkunde, KULeuven, Cel estijnenlaan 200B, B-3001 Leuven, Belgium andSterrenkundig Instituut ‘Anton Pannekoek’, Kruislaan 403, NL-1098 SJ Amsterdan, [email protected]. WaelkensInstituut voor Sterren k und e, KULeuven, Celestijnenlaan 200B, B-3001 Leuven, [email protected] study the time dependency of Vega-like excesses using infrared studiesobtained with the imaging photopolarimeter ISOPHOT on board of ISO. Wereview the different studies published on this issue, and critically check and reviseages and fr actional luminosities in the different samples. The conclusions ofour study differ significantly from those obtained by other authors (e.g. Ho llandet al. 1998; Spangler et al. 2001) who suggested that there is a glo bal power-law governing the amount of dust seen in debris disks as a function of time.Our investigations lead us to conclude that (i) for stars at most ag es, a large– 2 –spread in fractional luminosity occurs, but (ii) there a r e few very young starswith intermediate or small excesses; (iii) the maximum excess seen in stars of agiven age is about fd≈ 10−3, independent of time; and (iv) Vega-like excess ismore common in young stars than in old stars.Subject hea dings: (stars:) circumstellar matter—infrared: stars1. INTRODUCTIONThe detection of dust orbiting around Vega (Aumann et al. 1984) and other nearbymain-sequence stars by the IRAS satellite marks t he first detection of planetary material instellar systems other than our own solar system. In the years since then, much at tention hasbeen paid to understand the origin and evolution of this phenomenon, and on the questionif and how this discovery can tell us more about the planetary systems probably hidden inthose dust rings and clouds. One important aspect of the research was the study of temporalevolution of the amount of dust present in these systems. Already Backman et al. (1987)noticed from colour-colour studies, that systems like Vega, Fomalhaut and ǫ Eridani could bemore developed versions (with less dust) of the most f amous Vega-like star β Pictoris. Thissuspicion was further strengthened by submillimetre observations of t he four main membersof this class of objects. Submillimetre observations are very useful since they have a betterchance to measure the mass present in the system. Holland et al. (1998) showed that themeasured masses of the disks of several prominent Vega-like stars seem to follow a power-lawdependence with time. A stunning extra was t hat even the zodiacal dust cloud of the solarsystem seems to fit this trend.With the launch of the Infrared Space Observa t ory, ISO, in 1995, several research groupsset out to get a better handle on the time evolution of the Vega phenomenon. These groupsused similar observa t io ns like the ones obtained with IRAS, namely photometry at mid-IRto far-IR wavelengths to determine the dust masses in these systems. Several roads wherefollowed to improve the poor statistical significance of the IRAS studies: IRAS had shownthat some stars display the Vega-phenomenon, but the stars did not form a good samplein several ways. First, the ages of the stars were poorly determined. And second, thedetection limits of IRAS were such that large excesses could be detected, but for most starsthe detection of the photosphere of the stars was not possible. It was therefore easier toestablish that a star has a dust disk than to show that it does not have one.The first issue, concerning the a ge determination, was addressed by several groups (e.g.Spangler et al. 2001) by choosing stars in different clusters. Since the ages of clusters can be– 3 –determined rather accurately, this strategy should provide a much better calibration of thetime axis. The disadvantage of this approach was that most clusters are not very nearby,and the photosphere of most stars in the sample was out of reach also for ISO sensitivities.Habing et al. (1999, 2 001) and Silverstone (2000) addressed the second problem of theIRAS results on the presence of a debris disk. They selected volume-limited samples of starsin order to study the Vega phenomenon in field stars. In particular, they chose differentvolumes for stars with different spectral type (i.e. luminosity) in order to ensure that thephotosphere of the star would be detectable in all cases with good signal-to- noise. Thissample is t herefore mainly geared to check whether or not a disk is present. The obviousproblem with field stars is, of course, that it is much more difficult to determine accurateages for these stars, which may make a time-dust mass relation difficult to discover.After the launch of ISO, it turned out that the targeted photometric sensitivities weredifficult to r each. In fact, the limits could be reached, but only with a changed observationalstrategy (mini-maps) which required much more time than foreseen in the proposals. Thishas made the outcome of the studies less decisive than had been hoped. The task of cleaningup this question will be one of SIRTF’s major goals. Observers were faced with the decisionto either accept larger limits, or to observe fewer sources and/or at fewer wavelengths.Nevertheless, the groups working with ISO came up with age trends based on theirobservations. Studying the disk frequency around young A stars, Habing et al. (1999, 2001)concluded that there was a very clear break in the presence of a debris disk around 4 00 Myr.Most A stars younger than this limit showed Vega-like IR excesses while most A stars olderthan this limit did not show corresponding excesses. However, this work and the work ofDecin et al. (2000) also showed that several stars supposedly much older did also show IRexcesses.Spangler et al. (2001) published the results of two samples: stars in several clustersof various ages, and a number of field stars with less well determined ag es. In their data,Spangler et al. (2001) could determine a trend in the time evolution of the dust content ofstellar systems. They found a power-law dependence with a power fd∝ t−1.76. They furtherargued that a
View Full Document
Unlocking...