Large scale projection through the simulation called Illustris shows dark matter on the left transitioning into gas (ordinary matter) on the right. Image shows the universe on a cosmic scale with filaments spanning hundreds of millions of light years. A single galaxy is only a tiny blip on this image
New Delhi, May 7: Scientists have used supercomputers to simulate 13 billion years of the births and deaths of stars and the evolution of galaxies, creating a virtual universe that resembles the present-day observed cosmos more closely than ever achieved before.
The simulation by an international team of researchers spans a period 12 million years after the birth of the Universe in the Big Bang (13.7 billion years ago) till the present day, for the first time reproducing a realistic population of galaxies.
The success suggests that standard ideas about the evolution of the universe and the processes that shape galaxies are correct.
It is also consistent with longstanding ideas that 95 per cent of the universe remains unknown — a combination of enigmatic and invisible dark matter only experienced through its gravitational effects and an even more mysterious force called dark energy that is pulling the universe apart.
“This is exciting because our mock universe closely resembles the galaxies observed in the real universe,” Shy Genel, a post-doctoral researcher at the Harvard Smithsonian Center for Astrophysics in the US and team member, told The Telegraph.
The simulation study, led by Mark Vogelsberger at the Massachusetts Institute of Technology, appears tomorrow in the journal Nature. Earlier attempts to simulate the universe failed to create the mixed population of elliptical and spiral galaxies.
“This new exercise will allow astronomers to simultaneously model the universe on a really grand cosmic scale to the galactic scale — from a huge chunk of the observable universe to a single galaxy,” said Tirthankar Roy Choudhury, a scientist at the National Centre for Radio Astronomy, Pune, who was not associated with the study but specialises in the simulation of cosmic structures.
“Their predictions nicely match with observations,” Roy Choudhury told this newspaper. “The trouble with most previous simulation attempts was that they could either model large-scale structures of the universe or small galactic-scale structures — not both.”
Past simulations that tried to model the universe at the level of several hundreds of millions of light years could, for example, model the cosmos on its grandest scales, but were unable to model single galaxies. In such simulations, all galaxies looked the same.
Vogelsberger and his colleagues in academic institutions in Germany and the UK solved the problem through a combination of new rules to predict the behaviour of the cosmos on the large and smaller scales and using massive computing power.
Their simulation relied on several supercomputers and used 8,192 central processing units to perform calculations in parallel for three months. “This translates into an equivalent of about 2,200 years of non-stop computations on a single desktop computer,” Genel said.
Astronomers say such simulations are intended to help improve their understanding of galaxy formation and the two big unresolved mysteries — dark matter and dark energy.
“As our measurements of dark matter and dark energy get better and better, it will become more and more important to have accurate simulations that include all physical processes in galaxy formation,” Michael Boylan-Kolchin, a cosmologist at the University of Maryland in the US who was not associated with the study, told this newspaper.
“So, in addition to building better models of galaxy formation, simulations like these may also help us in our quest to understand the nature of dark energy.”