Tiger's Scientific Computing

Stellar dynamics in galactic nuclei (L. Subr, V. Karas)

Central galactic clusters are places of the largest stellar concentration. Stellar dynamics in these systems is influenced by many factors like gravity of the supermassive central black hole, mutual gravitational (encounters) or direct (collisions) interaction of stars or both gravitational and hydrodynamical interaction with gaseous environment (e.g. accretion disc).

Cores of distant galaxies are not observed with sufficient resolution and, therefore, only knowledge of integrated quantities is available. Their interpretation requires modelling of the whole system, which represents one of the most exhaustive computational tasks. In our simulations, we concentrate on influence of an accretion disc on the cluster structure. Ref.: L. Subr, V. Karas, J.-M. Hure, 2004, MNRAS, 354

Study of individual stellar trajectories is motivated mainly by recent observations of stellar proper motions in the centre of our Galaxy. We investigate migration theory of their origin. Ref.: L. Subr, V. Karas, 2005, Astronomy & Astrophysics, 433

Solar System dynamics (D. Vokrouhlicky, M. Broz, D. Capek)

Long-term evolution of planetary systems, including our own Solar System, is complicated mainly because of mutual gravitational interactions of planets. This necessarily results in extreme sensitivity on initial conditions known as deterministic chaos. An additional complexity occurs in the case of motion of small bodies (such as asteroids, comets or dust particles in rings or interplanetary region). This is because many forces of non-gravitational origin (such radiation pressure, solar wind drag or electromagnetic force) affect their dynamics on a long-term. A combination of chaoticity in the case of gravitational interactions and, typically, parameter-dependent modelling in the case of the non-gravitational effects means we need many different numerical simulations done before reaching quantitative conclusions.

Projects for which the Tiger cluster have been used:

  • Modelling of the late Miocene spike in He-3 record discovered in sea floor sediments of pelagic carbonate. We examine a possibility that this signal has been produced by massive Earth accretion of interplanetary dust from a disruption of a parent asteroid of the Veritas family. Ref.: K.A. Farley, D. Vokrouhlicky, W.F. Bottke, D. Nesvorny, 2005, Nature, in press.
  • Fitting the infrared emission of near-ecliptic dust bands. We developed a sophisticated code (SIRT) allowing us to predict infrared emission of dust particles released from specific sources (recent disruptions that have let to formation of Karin and Veritas asteroid families) and migrating toward the inner Solar system due to the Poynting-Robertson orbital drag. As an innovative feature, our model also accounts for the ongoing collisional griding of the dust-band population of particles. Ref.: D. Nesvorny, D. Vokrouhlicky, W.F. Bottke and M.V. Sykes, 2005, Icarus, in press.
  • Tumbling asteroids and meteoroids due to the thermal torques. It has been suggested that thermal torques can secularly drain angular momentum from rotation of small asteroids and meteoroids. Here we analyse how this process drives tumbling of the angular velocity vector in the body frame.

Reconstruction of physical parameters of asteroids (J. Durech)

Basic physical characteristic of asteroids (the rotation period, spin axis direction and shape) can be derived from their brightness time-variations. This lightcurve inversion method has led to about 100 asteroid models so far. The traditional lightcurves used for the inversion densely cover the asteroid's rotational phase during the night and require much observing time.

The completely new possibility is to use calibrated photometric data sparse in time. From the mathematical point of view, the time intervals between measured points is not a key factor for the lightcurve inversion. Photometric observations can be made in a more efficient way and a telescope can scan large area on the sky during the night. Contrary to the densely sampled lightcurves, the sparse data typically consist of a few brightness measurements per night.

The calibrated photometric data will be available from the Pan-STARRS project of the University of Hawaii in the near future (the first observations are planed for 2006). The amount of photometric data from this survey will be really huge compared to all previous projects and we expect tens of thousands asteroids model within the next decade - a number incomparable an unobtainable by any other observing technique.

At present we use the Tiger cluster for testing our inversion procedures on synthetic data or photometric data available from astrometric databases. Sparse photometry does not provide any information on asteroid's rotation period at first sight (contrary to the standard lightcurves) and that is the reason why the inversion process it much more time-consuming than in the case of dense photometric data. The period has to be searched within a wide interval (typically 2 - 20 hours) with a small step (typically seconds). For each trial period the best pole/shape solution is found using the Levenberg-Marquardt algorithm.

Last updated: 28.10.2005 (L.)