Free Astronomy Magazine July-August 2014

JULY-AUGUST 2014 COSMOLOGY All simulations undertaken so far have been influenced by this kind of limita- tions and shared the same inability to recreate the correct mix of elliptical and spiral galaxies that we can ob- serve in every epoch of the universe and in particular in that of our present day (z=0). In order to precisely recon- struct the evolutionary history of indi- vidual galaxies (and hence of the clusters and entire universe) it is indi- spensable to complicate the simula- tions by including several other properties in addition to the gravita- tional force. This entails combining to- gether several evolutionary models of galactic and sub-galactic size struc- tures into a single numerical process- ing step. In this way it will be possible, for example, to take into due consider- ation the influence had on the evolu- tion of galaxies by the supermassive black holes placed at their centre and the radiation emitted by them, the ex- plosions of supernovae, the super- winds from the star-forming regions and consequent distribution of free gas, the magnetic fields of different sizes and intensity associated to the different structures. In short, it is not sufficient to make a large number of virtual particles interact gravitation- ally (let us imagine them as 3D pixels), as it is necessary to also assign to those particles the typical properties of their real-life counterparts. From these premises, it can be easily gathered that by having available pow- L eft: a video of the universe according to Illustris. On the right, a large-scale view of the contempo- rary universe – cen- tred on the more mas- sive galaxies cluster – showing the transition between dark matter (bluish part) and gas (other colours). [Illus- tris Collaboration, NASA, PRACE, XSEDE, MIT, Harvard CfA] erful supercomputers, a very large number of virtual particles and more than reliable theoretical mod- els, it should be possible to simu- late the evolution of a portion of the universe that is representative of the whole universe, both on a small and large scale. But why start a simulation of that kind with all the difficulties and work it involves, when in the best of cases it can confirm the exis- tence and the physical and chemi- cal properties of already known structures? The answer is rather simple. Of those structures, in par- ticular galaxies and galaxy clus- ters, we only have snapshots that portray an almost insignificant moment of their existence. To get an accurate idea of how they have formed and evolve during billions of years, we can at best put in order of sequence corresponding subjects that supposedly represent different evolution stages of struc- tures that are similar to each other but placed in different epochs. It is like trying to understand the plot of a movie while having avail- able only a handful of frames ran- domly extracted from the reel. We can only speculate what might have happened between one frame and the next, but we cannot prove it. A specific simulation could, how- ever, reconstruct the missing parts by figuring out the entire plot from