Understanding galaxy formation is one of the most important and challenging problems in modern cosmology. In particular, star formation is perhaps the worst understood and least developed ingredients of modern cosmological simulations. This collaborative project, led by Dr Kravtsov, aims to reduce greatly the uncertainty in the physics of galaxy formation by developing new, observationally motivated models for the formation of stars and star clusters. Recent advances in numerical algorithms and computing power mean that cosmological hydrodynamical simulations can now resolve the actual sites of star formation in the interstellar medium, making it possible to track the formation and destruction of giant molecular clouds self-consistently. The planned suite of ultra high-resolution N-body and gas dynamic adaptive mesh refinement simulations will incorporate an unprecedented wealth of physical processes, including the formation of molecular hydrogen on dust and its self-shielding from ambient ultraviolet radiation, and self-consistent three-dimensional continuum radiative transfer. On smaller unresolved scales, the simulations will incorporate star formation recipes based on reliable empirical correlations. The planned research will study the molecular content of high-redshift galaxies, develop star formation prescriptions consistent with the observational correlations, and study star formation in different galactic environments. It will address the formation and evolution of massive star clusters, for which abundant observational data exist, and investigate the dependence of cluster formation efficiency on properties of the host galaxy.
This project is unique in its ability to bridge the large scales of galaxy formation and the small scales of star formation, thus developing links between those two research communities. By involving students, the work provides professional training of young computational cosmologists in cutting-edge simulation techniques. The results provide support for current and forthcoming space- and ground-based observing programs. The research results are tightly integrated into academic teaching, and public education and outreach activities. Three-dimensional visualizations of the simulations will be available for planetarium shows and public lectures, and also web-accessible.
This project focused on improvements in the way formation of stars isdone in computer simulations of galaxy formation. Most observationsthat inform us about details of galaxy evolution are observations ofstars. Therefore, modelling of how stars form in numerical models ofgalaxies is of paramount importance. As part of this project, PIKravtsov and co-I Gnedin have developed a novel model for modellingstar formation in such models. The model ties formation of stars withmolecular gas becayse regions of interstellar gas rich in molecules are known to be nurseries of young stars in real galaxies. To implement such model, the PI and co-I have developed andimplemented a novel method to track formation and destruction ofmolecules during the course of galaxy formation simulations. Thehydrogen molecules are fragile and easily destroyed by the intenseultraviolet light emitted from massive young stars, particularlyintense during the first three billion years in the evolution of theuniverse. In some galactic regions dark clouds, so–called because ofthe dust they contain, form a protective layer that protects thehydrogen molecules from the destructive light of other stars. Starformation in evolving galaxies thus depends on the intensity ofultraviolet radiation and amount of dust in interstellar gas. This model generically predicts that efficiency of star formation islower during early stages of galaxy formation, when there is littledust but the universe is filled with intense ultravioletradiation. Simulations done as part of this project showed that thisis indeed the case. This effect predicted by simulations have beenconfirmed by several observational studies by Arthur Wolfe, Hsiao-WenChen and collaborators. In the June 3, 2010 issue of Nature, RobertKennicutt, Plumian Professor of Astronomy and Experimental Philosophyand Director of the Institute of Astronomy at Cambridge University,noted about this work: "Gnedin and Kravtsov take a significant step inunifying these observations and simulations, and provide a primeillustration of the recent progress in the subject as a whole."
