Stars are the engines that drive how galaxies evolve. When massive stars die the resulting supernova explosions transfer energy to the gas in the galaxy and support it against gravitational collapse. Stars form all the elements heavier than Helium by nuclear fusion, and these elements are returned to the galaxy in the supernova explosions. Planets are formed as by-products of star-formation, and the energy that they emit may fuel life on those planets. An understanding of how stars form is crucial to understanding the evolution of galaxies since the Big Bang. We can test this process using revolutionary new telescopes like the Atacama Large Millimetre and sub-millimetre Array (ALMA) and the future Square Kilometre Array (SKA). These advanced facilities will make it possible to observe the different environments within other galaxies in unprecedented detail.

My research seeks to address these issues by using supercomputer simulations to create extremely detailed models of where gas is concentrated into clouds in different types of galaxies. These gas clouds are known as molecular clouds, because they mainly consist of hydrogen molecules, and are the stellar nurseries in which stars are born. I use the models to predict the temperature, density, chemical composition, and gas motions, and then investigate how easily the gas collapses to form stars, the type of stars formed, and the numbers of stars in the different clouds. This enables me to investigate whether molecular clouds and the stars formed in them are the same everywhere, or whether they vary with galactic environment. This is an important assumption as in observations we can only see the very brightest stars in other galaxies and have to assume that the fainter stars form in the same proportion as we see in nearby star-forming regions in our own Milky-Way Galaxy.

Previous models of star-formation always started with simple estimates of how clouds might begin, but uniquely, my collaborators and I use actual galaxy models. A major innovation is the use of a chemical model to predict the composition of the gas, which will determine the emission seen by a telescope. For example, ALMA is sensitive to the emission from carbon monoxide gas, and the SKA will be sensitive to emission from atomic hydrogen. Using the computer simulations generated from this work we are aiming to make synthetic observations of the molecular clouds in other galaxies. In other galaxies we know that the amount of gas, the structure of the galaxy, and the chemical composition may be different, and so how clouds appear when observed will also be different. We can compare the computer generated emission maps to the observed maps as a reference guide to determine what the properties of the observed clouds actually are. Astronomical observations are 2D projected images of what are 3D structures, moreover, the emitted light changes with the temperature and density of the gas. This makes molecular clouds hard to understand without a good model to compare against. This is particularly true of modern telescopes as due to their improved resolution and sensitivity the observations are very complex due to the fine details that they can see. Consequently, the models from our studies will be crucial in using ALMA and the SKA to determine how stars form throughout our Universe.