The "Stars and Planets" group was one of three groups comprising
the Edinburgh workshop. At an early stage the main group divided
itself into two, allowing those whose science interests are entirely focused
on star formation problems to pursue that very obvious avenue .
The second group, which I report on here, tried to take a broader
perspective across the whole of stellar astrophysics, in order that nothing
was missed there. We were Penny Sackett, Dainis Dravins, Pat Roche and
I (Janet Drew) as I recall (with Adrian Russell present much of the time).
Accretion disks are commonplace in the Universe and exist on a range
of spatial scales: active nuclei in galaxies are believed to be fed by
matter transported through them; accretion disks lie at the heart of the
star and planet formation processes and become important again in the late
stages of stellar evolution in a range of interacting binaries. Most recently,
the formation of an accretion disk has been mooted as part of the mechanism
for creating `hypernovae'.
A long-standing problem has been the origin of the viscous shear that is central to the inward migration of matter through an accretion disk, and the outward transfer of angular momentum. It was recognised over two decades ago that `normal' viscosity on the molecular level was orders of magnitude too weak to explain the timescales observed in astrophysical disks. Currently there is much interest in the role that turbulent MHD processes may play in creating this `anomalous' viscosity. But there are problems within even this concept (e.g. how to sustain the viscosity in the absence of significant gas ionization) and it is clear that the work is cut out for the astronomical community to devise observational tests of this model and subsequent variants. What this amounts to is that observations are needed that can characterise the interior structure of accretion disks.
Consider the case of a disk around a compact object such as a white dwarf or neutron star (objects of mass of order 1 Msun). Currently feasible spectroscopy on 4 -- 10-metre telescopes using the integrated light from such systems has failed to teach us much at all about the interior structure of such disks. Partly this is for technical theoretical reasons (realistic modelling is still beyond us). But it is also because existing data only tell us about grossly time-averaged spectral properties and there is also the fundamental limitation that much of the emergent light may only originate in superficial layers that tell us little. To do better than this, and really prise open the interior workings of accretion, a proper exploration of the time domain is needed. The relevant timescales for accreting compact objects are short: the period of the innermost Keplerian orbit around a 1-Msun white dwarf is about 10 secs. It is plausible that phenomena associated with the turbulence implicated in the viscosity exhibit timescales this short or even shorter. White light observations of flickering have already detected timescales in this regime. What would we find if we could take spectra, well-enough resolved in wavelength to reveal the details of spectral line shapes, at a rate of 1 per second? It might then be possible to construct and analyse lightcurves for specific doppler shifted wavelengths within spectral lion profiles (specific locations within the accretion disk observed, in effect). Data of this nature amounts to the much-needed window onto the dynamic processes we expect to be at work in such environments.
A related class of time domain observations would be reverberation observations of a more refined kind than are already undertaken on today's telescopes. Light travel times across WD/NS/BH interacting binaries are on the order of 1 sec.
The accretion disk in star and planet formation is a problem that captures
most imaginations much more vividly. The analogous experiment to the compact-binary
time-resolved spectroscopy, in this setting, could be carried out in a
directly spatially-resolved fashion with an ELT. Since the disks of young
stars are absolutely much larger, the timescales of interest will be correspondingly
longer. The problem will be more complex in that the disk constituents
are a mixture of gas, dust and larger rocks. The reader is referred to
other submissions that will deal with the quantitative issues appropriate
here.
Figures to calculate on: V mags in the range 12 to 20, to be observed at a sampling rate of 1 sec or better, at a spectral resolving power of 5000-10000.
The optical is needed in order to emphasise the accretion process rather than any companion object. (In the best of all possible worlds one would probably most like to access the space UV!)
For the study of disks around young stars, spatial resolution down to 1 mas is desirable at least at IR wavelengths. Coronography is also required to suppress the light of the central star. This last is something a wide range of stellar astrophysics applications will need - hence some hard data on the degree of nulling achievable will be useful.