Name: Cristiano Sabiu
Institution: University of Portsmouth
Research: Distribution of galaxies in the universe
Cristiano Sabiu is a recent PhD graduate from the University of Portsmouth where he studied the distribution of galaxies in the universe under the supervision of Prof. Robert Nichol. The distribution of galaxies can tell us a great deal about the underlying physics which govern the dynamics and evolution of the universe.
There are currently two unsolved issues in cosmology. It is known from recent observations that the universe is not only expanding but is doing so at an accelerated rate, but what is driving this? This unknown force is usually called dark energy. The second issue is that it is known from observations that an extra source of gravity must exist in the universe. It is thought that there are particles which do not emit light and only interact through the gravitational force, this is usually called dark matter.
It is thought that galaxies form where there is a concentration of dark matter but the exact relationship between the dark matter and galaxies is hotly debated. The relationship is usually characterised by a very simple function known as `bias'. Cristiano used an alternative to the standard bias, known as the Halo Occupation Distribution (HOD). The HOD is a method for placing galaxies within a Dark Matter distribution. Although the HOD has many advantages over the bias prescription, its exact form is not well understood.
His research focused on finding the exact form of the HOD, mathematically this involves solving for the unknown variables in an equation but in reality he is exploring how the galaxies we see are connected to the invisible dark matter. This required running many large scale N-body simulations of dark matter. From these simulations he created a multiverse of almost 2000 mock universes which explored multiple HOD models. These were then compared to the actual galaxy distribution as observed by the Sloan Digital Sky Survey (SDSS) allowing the placing of tight constraints on the model parameters.
The simulations were run using the publicly available GADGET2 code. GADGET2 is an N-body integration routine which can run cosmological simulations on massively parallel computers with distributed memory. GADGET2 uses an explicit communication model that is implemented with the standardised MPI communication interface. The code can be run on essentially all supercomputer systems presently in use, including clusters of workstations or individual PCs.
Cristiano simulated the large scale structure using over 100 million dark matter particles and, as his models were generally very spatially homogeneous, the code showed excellent scalability. He completed a series of cosmological simulations, using a 128 CPU configuration on the NGS-RAL facility and successfully ran 20 full scale simulations, requiring ~100,000 CPU hours over a 12 month time period.
Cristiano said “GADGET2 was installed for me on the NGS and optimised for their system. It is now available to all NGS users so if you fancy creating your very own Universe why not give it a go!” adding that “Without the NGS my PhD project could have taken 10 years!”.
Project funding - STFC
Download a summary slide of this case study.
