IAU Logo

IAU Symposium 239

Convection in Astrophysics

21 - 25 August 2006
Prague, Czech Republic

IAU General
Assembly 2006 Logo

Scientific Rationale

Convection is ubiquitous throughout the Universe. Its main physical consequences are heat transport, mixing, interaction with mean flow and magnetic fields and the dynamo generation of magnetic fields.

Convection occurs in the majority of stars including the sun. During the last three decades convection has become perhaps the largest factor of uncertainty in the physical modeling of many problems in stellar evolution, stellar structure, pulsational stability and model stellar atmospheres, to name just a few. Nevertheless, considerable progress has been achieved in our understanding of solar and stellar convection through new types of observations. These include high resolution spectra and time series of spectra of solar granules, as well as accurate measurements of the depth of the solar convection zone by means of helioseismology and of the properties of subsurface convection by local helioseismology.

At the same time, numerical simulations of solar and stellar convection have become more and more powerful because of enhanced numerical resolution and more realistic microphysics. Nevertheless, modeling and theoretical understanding of how convection interacts with other physical mechanisms is still very much wanting, as is the successful application of knowledge gained on convection to the unresolved problems of solar and stellar astrophysics.

Convection is also common in planets sufficiently large to keep an atmosphere of their own. In such atmospheres convection is usually a local phenomenon of limited horizontal extent. Stellar convection for comparison usually occurs as a global instability within a shell or a sphere. Convection is also present inside planets which have a heat source in their interior and where it can occur as a global phenomenon very much like in stars. The interior of Jupiter provides one such example. Planetary research on flow dynamics and thermal properties benefits from the possibility of in situ data obtained from probes which is not available to solar and stellar research. On the other hand, the limitations to computational capabilities and the difficulties in the modeling and the theoretical understanding of the interaction between convection and other physical processes are very similar to those encountered in solar and stellar research.

The conference will deal with theoretical and observational aspects of convection in a wide variety of objects from our own Solar system and beyond, in the Sun, stars and planets. The sessions will be organized around physical processes with which convection interacts in the objects studied, whether it be gaseous giants, a brown dwarf, the sun, or other stars.


The sequence of sessions and the basic range of topics will be as follows:
A - Modelling of convection and radiative transfer
B - Observational probes of convection
C - Convection in planets and brown dwarfs
D - Stellar evolution, nucleosynthesis, and convective mixing
E - Oscillations, mass loss, and convection
F - Convection and rotation
G - MHD convection and dynamos
The following aspects will be covered by this programme:

Modeling of and numerical simulation techniques for convection

Convection has been posed as a serious and difficult problem in astrophysics for decades. The initial, and conventional, approach to handle convection was based on heuristic arguments. The physical understanding and applications were both irked by uncertainties. In the past two decades, numerical simulations have made very important contributions to the understanding of astrophysical convection. The extent to which numerical results are applicable and reliable, however, is not without queries. In recent years, alongside the achievements made by numerical simulations, considerable development has also been made on the theoretical side, in attempting to improve the mathematical treatment of the convective turbulence. Contrasting of results from mathematical modeling, numerical simulation, and the conventional approach have also been made. The study of astrophysical convection has become a big field, the spectra of approaches, techniques, and phenomena involved are wide.

Observing convection

Until recently convection and turbulence have been treated in rather ad-hoc manner with adjustable free parameters within model atmosphere and interior codes. As such, these can be selected to agree with observations without any regard to their true physical meaning, even to the extent of being un-physical. The new theoretical developments to be discussed at this conference enable true physics-based predictions to be tested against observations in order to improve our overall understanding of the physical mechanisms. The increased sophistication and precision of instruments is pushing our theoretical models to their limits and beyond. Example include high-resolution measurements of absorption line profiles and line shifts as diagnostics of atmospheric convective flows, helio- (and astero-) seismological probing of internal convection zones, direct imaging of surface convection of the sun and other stars (through interferometry), and direct observations of atmospheric circulations within planets of the Solar System.

Convection and radiation

The most outstanding property of solar surface convection, and of other stars with surface convection, is its strong interaction with the radiation emitted into space. Theoretical modeling of this process has remained extremely difficult and is one of the unresolved problems in the theory of stellar atmospheres. Numerical simulations of solar and stellar surface convection at the same time have made significant advances and in many cases agree so well with the data that they now become an excellent template for models and for understanding the physical processes of radiation-convection interaction. Despite the enormous success in the solar case, the simulation of different types of stellar atmospheres has not been without surprises, due to the complications involved by non-local thermal equilibrium in stellar atmospheres, or interaction with other physical mechanisms such as rotation, magnetic field, uncertainties in opacities or the circumstellar environment, and others..

Convection in planets and brown dwarfs

The interior of gaseous giants is expected to transport its heat outwards through deep convection zones. The same region is also considered to give rise to planetary dynamos. The close similarity of these physical scenarios to stellar counterparts makes a review of the state of our knowledge and the possibility of exchange with the solar and stellar physics communities particularly interesting. Deep convection zones are expected to appear already during formation stages of gaseous giants, brown dwarfs, and low mass stars. Its effects are observable through the depletion of Lithium in the latter, while brown dwarfs show abundances close to those of the interstellar medium.

Convection and nuclear reactions

The coupling of convection and nuclear reactions is recognized as an important process in stellar structure. It obviously requires consideration in the dynamical phases (supernovae nucleosynthesis), but also during the fast phases of the post-Asymptotic Giant Branch (AGB), such as Sakurai's object, when 'reactive' convection can occur. It is also crucial for the expected nucleosynthesis when convection reaches the hydrogen burning shell in the high luminosity phases of massive AGB evolution, and for determining the properties of excursions through the blue loops (and thus the Cepheid phase) during helium burning in the cores of intermediate mass stars.

Convective mixing

Convective mixing at the borders of convective cores (either as 'overshooting' or, in the convective helium burning cores, possibly as 'semiconvection') has been recognized since the seventies as one of the key processes to understand the basics of stellar evolution, e.g. the lifetimes and width of the main sequence band, but the physics of such overshooting is poorly understood and consequently is crudely parametrized when included in the construction of stellar models. Progress in this area is perhaps one of the most needed in stellar astrophysics, as convective mixing also influences later and final stages of stellar evolution. Metallic fingers (thermohaline convection) have been suggested to explain chemical peculiarities observed in stars (as the final consequences of binary interaction or recently absorbed close-by planets).

The question of double-diffusive convection has come up as an explanation for the unexpected properties of the progenitor of SN 1987A and the existing stellar evolution calculations are still very much wanting in that respect (these have to be considered together with the binary scenario proposed for that supernova as well). No combined treatment of rotation and double-diffusion has ever been implemented in stellar evolution calculations of massive stars. Progenitor models of core-collapse supernovae hence still suffer a considerable amount of uncertainty caused by a very incomplete modeling of convective mixing.

Convection, oscillations and mass loss

Turbulent convection can stochastically excite stellar oscillations and the properties of the oscillations depend on the time dependent interaction of the oscillations with the convection. The measured properties of the oscillations provide diagnostics of both the global structure of a star (depth of convective zone, radius of convective core, overshooting), models of turbulence and, in the solar case, diagnostics of the properties of the subsurface convective layers. The behaviour of oscillating stars (Cepheids, delta Scutis, Miras) and the red edge of the zone of instability strip in the H-R diagram - depend on the detailed properties of the interaction of the pulsation with convection. In late type giants this may also drive mass loss and envelope ejection. Understanding of all these topics is currently being advanced by theoretical modelling, numerical simulation and observation, but there is a long way to go before we can be confident that we understand and can model convection-pulsation interaction.

Convection, rotation, and disks

In astrophysical settings, rotation is not less ubiquitous than convection, and the coupling of the two provides the drives for many energetic astrophysical phenomena. The most basic among them, differential rotation, is still a hot subject being intensely pursued. The relative importance of local and global processes is a question under heated debate. At any rate, in the solar/stellar community, since differential rotation is known to vanish below the solar convection zone (through helioseismology), it is generally accepted that differential rotation is a consequence of the interaction of rotation and convection. In the planetary community, on the other hand, the claim that such interaction is responsible for the wind bands of the outer planets is not as popular.

Rotation convection and turbulence are important in astrophysical disks too, be they accretion disks around compact objects or proto-planetary disks around young stars. Our understanding of them has been enhanced by complex 3-D hydrodynamical simulations.

MHD Convection and dynamos

The magnetic fields in stars and planets (and in many other astrophysical objects) are believed to be generated by dynamo action of convection. Understanding the interaction of convection and magnetic fields is therefore of major importance for astrophysics. Under what conditions is a small seed field amplified? How can turbulent convection produce a mean field on a scale much larger than the scale of the turbulence? Can a strong magnetic field inhibit convection? Can we model the solar dynamo? Can a dynamo operate in stellar convective cores? Is rotation a necessary ingredient for a dynamo to operate, and how do rotation and magnetic fields effect the transport of energy in convective regions? These are very difficult problems but progress has been made in 3-D simulations of MHD convection and of rotating magnetic envelopes and cores, and in mean field models of the solar and stellar dynamos

Contact IAU Symposium 239