ON A LARGE SCALE

Course: Analytical and numerical models of galaxy formation

Prof. Simon White
Max-Planck Institut für Astrophysik
GERMANY

Nearly 90% of the mass contained in big galaxy clusters is in some unobservable form; dark matter is a key element for models explaining the structure of the Universe. But the hot intergalactic gas discovered by X-ray satellites in the 80's fails to provide sufficient dark matter to solve the problem. In this interview, Simon White, Director to the Max-Planck Institute for Astronomy, Munich (Germany), and one of the proposers of the cold dark matter cosmological model (CDM), the currently standard model explaining the large-scale formation of structure in the Universe, points out the most important successes and shortfalls of this model.

A simulated region of the Universe at the present time.

What are the main successes of CDM in explaining the recent data on the evolution of galaxies?

"In this field I think the main influence of hierarchical cosmogonies like CDM has been to teach us to see galaxy formation as a process rather than an event. In the past there was a tendency to think that galaxies of different types are like animals of different species, each born at a well-defined moment and following a standard type-specific path to maturity. In hierarchical models galaxies don’t have a definite date of birth; their stars form throughout cosmic history, and individual galaxies are constantly changing their identities and their structure. Thus high redshift irregular galaxies can merge and make present-day spiral bulges, spirals can merge to make ultraluminous starbursts and perhaps quasars before settling down as ellipticals, ellipticals can grow new disks and become spirals again. Recent data show clearly that the galaxy population at early times differed in many ways from the one we see around us today, that young galaxies were typically smaller, more irregular, more gas-rich, and more active than they are now. Birth and transfiguration are occurring throughout the wide span of cosmic history that we can now observe directly. CDM-based models are useful not only because some variant may, perhaps, turn out to be the correct theory for the growth of structure, but also because they suggest ways to characterise the dynamic processes of galaxy growth and transformation."

What are the main shortfalls? Can we ‘fix’ them?

"Many aspects of our picture of galaxy formation remain incomplete, and current CDM-based models certainly fail to reproduce some aspects of the data. For example, the structure they predict for the dark matter haloes of individual galaxies appears inconsistent with rotation curve data for dwarf and low surface brightness galaxies. In addition, current simulations of spiral galaxy formation in CDM-based models are unable to produce galaxies with disks as large as those observed. It is unclear whether these difficulties reflect a fundamental problem with the CDM model itself, or are a consequence of our lack of understanding of many of the complex physical processes (most notably those associated with the birth and death of stars) which regulate the formation and evolution of galaxies."

Galaxy evolution is measured by comparing local and distant samples. Do we know the properties of the local galaxy population sufficiently well?

"I think not. It is surprising, but in many ways we have better statistical data on the properties of the distant galaxy population than we do for nearby galaxies. For example, there is still no reliable determination of the global disk-to-bulge ratio, the mass of stars in the local Universe which lie in the disks of irregular, spiral and S0 galaxies, as compared with the mass in spiral bulges and elliptical galaxies. We badly need to get distributions of the fundamental galactic parameters (masses, sizes, luminosities, gas fractions, star formation rates, metallicities and characteristic velocities for both bulge and disk) for a truly representative sample of galaxies in the low-redshift Universe. This is quite feasible with current instrumental techniques, but it is a lot of work and unfortunately it is not seen as a glamorous enough project to feature high on most priority lists."

The theoretical views on galaxy formation are largely based on large numerical simulations. As the speed delivered by computers increases, models may deliver more physics and finer resolution. But simulations are deterministic. Do we need to introduce processes such as chaos and feedback in the simulations of growth of structure?

"Many aspects of current simulations of galaxy formation are chaotic in the sense that detailed results are extremely sensitive to initial conditions but statistical properties show convergence towards apparently "universal" behaviour. An interesting example which I have worked on myself in recent years concerns the structure of dark matter haloes. These are found to exhibit very similar average radial density profiles for all masses and in all variants of hierarchical cluster formation theories like CDM. On the other hand, it is certainly true that there are many processes which our current simulations do not represent even approximately correctly. The most important of these relate to the structure of interstellar gas, the way this gas turns into stars, and the effects that the stars have on the gas (the "feedback") as they age and die. I do not think any foreseeable improvement in computing power will allow us to simulate these processes reliably and in detail."

PROFILE

Born in Kent (England), 1951, SIMON D. M. WHITE has spent most of his life overseas. Linked to different American universities and scientific institutions between 1977 and 1994, he is currently Director of the Max-Planck Institute for Astronomy in Garching, Munich (Germany).

White was awarded his PhD in Astronomy at the University of Cambridge (UK, 1977), with a thesis on 'The Clustering of Galaxies', conducted under the supervision of Prof. Donald Lynden-Bell. The ideas then suggested about dark matter distribution in the intergalactic space were confirmed when the 'Einstein' X-ray satellite started to send detailed images of galaxy clusters.

As an expert in large-scale structure and formation of galaxies, White was a member of the Hubble Space Telescope Working Group on Galaxies (1984-85) and was part of the Time Allocation Committee for that telescope (1989). From 1991 to 1994 was a member of the Scientific Advisor Committee to the Gemini Project.

Together with Sir Martin Rees (also a participant, with Prof. Lynden-Bell, at the V Canary Islands Winter School of Astrophysics), White proposed in 1978 a model, still valid today, for the formation of galaxies, in which they held that galaxies formed from remnant gas kept in the gravitational potential wells of a population of dark halos in the process of fusion.

Noticeable in White's research activities are also his contributions to the theory of a cold dark matter Universe as a current standard model for the formation of structure. The technique set by the 'Gang of Four', to which White belonged, has become the usual method for comparing those models with observational data.

Member to the Max-Planck Society since 1995. The American Astronomical Society (Helen B. Warner Prize, 1986) and the British Royal Society (member, 1997), have also aknowledged his contribution to Astronomy.