The origin of galaxies and large-scale structure

The origin of galaxies and other large-scale structure in the universe remains an outstanding cosmological enigma. Why, given that the universe was so smooth and homogeneous at early times, were there tiny fluctuations which later collapsed gravitationally to form the structures we observe today? What caused these primordial fluctuations? How can we reliably describe their nonlinear evolution to make quantitative predictions to compare with present day observations of galaxy distributions? If you want to learn some basics about galaxies refer to:

At present there are two over-riding and competing paradigms for the origin of this large-scale structure:

1. Inflationary models. A period of extremely rapid expansion in the early universe - known as inflation - solves quite a number of cosmic enigmas. The key point here is that during inflation small quantum fluctuations get amplified up to enormous cosmological lengthscales - this is a manifestation of Heisenberg's uncertainty principle combined with the rapid expansion. These fluctuations later grow to form galaxies.

2. Cosmic defect models. Supermassive topological defects, like cosmic strings and textures, can form at cosmological phase transitions in the early universe. The subsequent defect network evolution is complex and nonlinear and it can have dramatic gravitational effects. The defects form the seeds around which galaxies form.

One of the aims of the UK-CCC is to examine the quantitative predictions of these two competing paradigms to determine which, if either, is consistent with the rapidly improving observational data describing the distribution of matter in our universe. Each of these paradigms has a wide variety of possible scenarios with different underlying models of the early universe which, in turn, are affected by different cosmological parameters in the late universe. One key goal is to differentiate between these possible models by direct comparison with the empirical data.

Another key goal is to develop methods to reliably elucidate the nonlinear processes which give rise to the observed properties of galaxies and clusters of galaxies. This involves including complex hydrodynamic effects of gas in numerical codes describing gravitational collapse of matter in an expanding universe.

The Virgo project

Computer simulations play an essential role in comparing the predictions of theories for the content and early evolution of the universe with observation and have become the primary method for testing such theories against astronomical data. Many of the basic concepts and simulation techniques were pioneered by UK astronomers in the 1980s. Our leading position was lost to US competitors who enjoyed superior computing power in the early 1990s and has only recently been restored through the work of the Virgo consortium. This international collaboration has the largest programme of cosmological simulations in the UK and one of the largest in the world. The recent `best of breed' Cray T3E/Origin 2000 purchase represents a small step towards the upgrade in computing resources essential for this programme to remain internationally competitive.

Simulations of large-scale structure

The first, still ongoing, project of the Virgo consortium consists of simulations of large volumes selected from all currently popular cosmological models. These include high resolution simulations with both dark matter and gas, being the first ever to resolve individual galactic halos and individual gaseous ``galaxies". The matter distribution in a 17-million particle simulation performed on a Cray supercomputer is illustrated below.

The scientific issues that will be addressed with these and future simulations include:

The structure and clustering evolution of the dark matter. Virgo simulations will provide the most accurate measurement so far of the evolution of the clustering properties of the dark matter in different cosmological models, over a wide-range of lengthscales.

The clustering evolution of galaxies. Quantifying the distribution of dark matter halos is an intermediate step towards a major goal of modern cosmology: understanding the distribution of galaxies themselves. The dark matter plus gas simulations will allow us to actually study ``galaxies'', identified as dense knots of cold gas associated with dark matter halos.

Gravitational lensing. Large mass concentrations act as gravitational lenses, amplifying and distorting images of background galaxies. Gravitational lensing thus provides a direct and elegant means for mapping the dark matter in clusters and Virgo simulations have provided the best to date for predicting the expected outcome of gravitational lensing observations in different cosmologies.

The dynamics and X-ray properties of galaxy clusters. Current dark matter plus gas simulations typically form one rich cluster per computational volume. With the ability to resolve individual galaxies within clusters, we can address a number of important unresolved issues.

Structure formation with topological defects. Viable paradigms for structure formation in cosmic defect models present considerably more formidable computational difficulties than their inflationary counterparts. In a collaborative effort between particle cosmologists and VIRGO, we wish to make more detailed quantitative studies of structure formation in these models (repeating the programme described above).

Simulations of galaxy formation

In the standard cosmological framework, the gravitational amplification of small primordial density perturbations leads to the formation of bound clumps of dark matter by the process of hierarchical clustering. Cooling is very efficient on galactic and subgalactic scales and, in the absence of heating sources, gas is expected to quickly lose its pressure support and collapse to the centre of virialized halos where it eventually turns into stars. The non-linear nature of the mass distribution on small scales and the complexity of hydrodynamic, radiative and stellar processes make supercomputer simulations essential for constructing realistic models of galaxy formation. The issues that we plan to address include:

The properties of protogalaxies. The Hubble Space Telescope is now providing quantitative information on properties of galaxies at intermediate and high redshift such as their morphologies, sizes, colours, and star formation rates. Virgo simulations will help establish the conditions required for this evolution to occur and the processes that determine the observable properties of high redshift galaxies.

The visible effects of galaxy mergers. The observational community is beginning to accept the view, long held by theorists, that mergers are the key process that drives galaxy formation. The simulations that we plan to perform will quantify the merger rates of galaxies in different cosmologies, the signatures of recent mergers, and the effect of mergers on galaxy morphology.

The angular momentum of galactic disks. Although perhaps not widely appreciated, the origin of galactic angular momentum remains a major unsolved problem. We plan a series of simulations specifically designed to test the hypothesis that it arises from feedback mechanisms that prevent all the gas condensing into dense fragments.

Primeval gas clouds. One of our high priority programmes is a study of the properties of the Lyman-alpha clouds seen along-the-line of sight to quasars. For this purpose we plan to carry out gas plus dark matter simulations similar to those described above, but in much smaller boxes. This programme will address the relationship between the Lyman-alpha clouds and the dark matter distribution, as well as other issues.

Further information

For a more detailed discussion refer to the Virgo consortium home pages or the original UK-CCC scientific case: