Population III stars were not only smaller than believed, they actually formed in binary systems, that is, pairs of stars that orbit a common center, say the results of a new simulation.
"For a long time the common wisdom was that these Population III stars formed alone," said Brian O'Shea, a Michigan State University assistant professor of physics and astronomy who did the research with two colleagues. "Researchers also have believed that these stars were incredibly massive – up to 300 times the size of our own sun. Unfortunately, the observations just didn't jibe with the simulations we created."
Another clue was that so-called "metals" – all elements other than hydrogen and helium – that are now found in newer stars don't necessarily match what was thought to be produced when the very massive first stars died.
Computer-simulated, false-color image of the formation of a binary stellar system. Photo Credit: Kavli Institute for Particle Astrophysics, Stanford University
Very old Population III stars were made essentially of hydrogen and helium. As the stars aged and exploded as supernovae, other elements were formed and these "metals" began showing up in newer stars.
"What we have here," said O'Shea, "is a fundamental disconnect between observations and theory, because these really massive stars would have produced a different set of metal abundances than what we see in old stars in our galaxy. If a lot of the Population III stars end up being in binary systems, then overall they would be less massive and so when they inevitably died, the metals they produced would be in much better agreement with what we see observationally."
O'Shea and his colleagues are theoretical astrophysicists, as opposed to traditional observational astronomers. They use supercomputers and custom-designed software to study the formation of cosmological structures such as galaxies.
What really drove this work, O'Shea said, was the development of faster, more powerful computers.
Computer-simulated image of a binary Population III star system forming approximately 13 billion years ago. Visualization by Ralf Kaehler, Matthew Turk and Tom Abel.
"All of the earlier simulations suggested that when these stars formed they were single, massive stars," he said. "But now we have faster computers, and we're able to work out our models to a higher level of detail. Our new simulations found that when you actually resolve everything it is possible that once the gas was going to make the stars, it was bound together tightly enough to make binary stars."
Did they give any details on the set of assumptions employed in setting up their simulations? Presumably the Standard (Cold Dark Matter) Model?
I ask because I've more than a passing interest in the non-standard models of space and gravity: bimetric, MoND, TeVeS, all of which challenge some of the basic assumptions of general relativity (specifically, collapsing the gravitational and geometrical tensors into a single metric), without losing their empirical viability.