Simulating the evolution of the early Universe on computers is the starting point for cosmologists' understanding of structure formation in the cosmos. With techniques to pursue both a large volume of simulated universe and high spatial resolution, KIPAC researchers are leading the charge against one of the foremost computational challenges in astrophysics.
Multi-scale density field used as an initial condition for simulations. To achieve large volume and high resolution, the density perturbations are resolved at significantly higher resolution in the central region and only coarsely sampled in the exterior.
Cosmological simulations have become an invaluable tool to study the formation of cosmic structures over an enormous range of scales, crucial aspects of the evolution of the Universe which cannot be directly observed. Starting from the tiny density perturbations that also give rise to the observed temperature fluctuations in the cosmic microwave background radiation, simulations with computers - where millions of particles are followed according to the known physics and starting conditions - are used to study the everything from the formation of the first stars in the Universe at scales of a few solar radii, the formation of galaxies at scales of several thousand light years, up to the formation of clusters of galaxies on million lightyear scales, and even the large-scale structure of the Universe on scales of a billion light years.
A major challenge in these cosmological simulations is to bridge the gap between cosmic sample variance - which requires a large volume to be simulated lest an atypical region be observed - and sufficient resolution in the single structures of interest to study them in detail. To meet this challenge, KIPAC postdoc Oliver Hahn and Professor Tom Abel have developed a new computer code named 'MUSIC' to generate multi-scale 'adaptive mesh' initial conditions which allows simulators to face this challenge and enable the study of study cosmic structure formation at very high local resolution over a large volume.
In the adaptive mesh technique, the MUSIC code generates a high-resolution realization of the density field in the early universe in a small region of space, while the density field outside of this region of interest is sampled at much coarser resolution, while still allowing a full accounting of any effects on more densely sampled regions of interest. Such multi-scale initial conditions allow simulations to focus computational resources in the subsequent self-gravitating evolution of the primordial fluid of baryonic gas and dark matter on the region where an object of interest - be it a first star, a galaxy or a galaxy cluster - will form.
Compared to previous approaches, MUSIC produces multi-scale initial conditions with significantly reduced errors - typically two orders of magnitude lower - and runs with a very low memory footprint enabling future simulations with greatly increased resolution and accuracy in the initial conditions. This will allow researchers to face the challenges of simulating both precision cosmology and high-resolution astrophysics of the formation of cosmic objects such as galaxies and galaxy clusters. Understanding their formation and evolution is of key importance for current and future observational efforts that aim at measuring the effect of dark energy on cosmic evolution and thereby reveal its nature. Furthermore, studying the formation of galaxies with increasingly sophisticated and computationally demanding simulations enables researchers to study the properties of the supermassive black holes that they host at their centers, thus probing regions where the extremes of high energy physics are unobtainable by experiment.
This work is based on a paper to be published in Monthly Notices of the Royal Astronomical Society and available from astro-ph at arXiv:1103.6031.
Tidbit author: Oliver Hahn and Jack Singal