A new study of the Cosmic Infrared Background claims to have discovered evidence for the oldest generation of stars - the so-called POP III stars (older than the POP I and POP II generations, which one sees directly in the sky). The work, by Alexander Kashlinksky and collaborators, is described in an article in this week's Nature, and is also the subject of a Nature News piece. There is also a related article on the BBC News website. From observations of younger stars in the universe, and from indirect evidence from measurements of the Cosmic Microwave Background radiation, it has been suspected for a while that there existed an early generation of stars (born less than 200 million years after the big bang). These stars are needed to fuse hydrogen and helium into some of the slightly heavier elements (that astronomers refer to as "metals"), which are seen in the later generations of stars. Further expectations for the existence of these objects come from large scale computer simulations of structure formation in the universe, which, starting from initial conditions predicted from cosmic inflation, see early structure which leads to highly clustered groups of extremely massive POP III stars. Kashlinsky and collaborators, using NASA's Spitzer Space Telescope (the Spitzer project also has a press release about this) are claiming not just evidence for the existence of POP III stars, but also that their results imply the high mass and clustering properties predicted by the simulations. Kashlinksky is quoted on the BBC website as saying
"It seems these first stars were quite unlike those we see today. They were huge thermonuclear furnaces; few and far between, but they burned ferociously because they were so massive"
Searching for these early stars isn't like looking for the stars one sees in the night sky. POP III stars would be so old and dim that one needs to infer their existence from the imprint they leave on the Cosmic Infrared Background (CIB) - the bath of infrared radiation that fills the universe. The hard part about this is that pretty much every other glowing object in the universe also contributes to the CIB, masking and distorting the signal from the earlier POP III stars. Removing these "backgrounds" to tease out the POP III signal is no easy task, and indeed, Kashlinsky and collaborators have taken a year to clean up the Spitzer data and arrive at their result. Given all the sources of noise in the CIB, what is surprising to some astronomers is how large and clean a signal these authors have found at the end of the day. Richard Ellis, a well-known astronomer from Caltech, comments in the Nature News article
"When people have looked for these signals before they found them to be barely in the realm of detectability, but this data looks really clean, ... I'm very surprised that the signal is so strong."
and on the BBC he adds
"... even a minor blunder in removing these foreground signals might lead to a spurious result ... A number of untested assumptions involved in allowing for unobserved galaxies could represent a weakness in the analysis."
However, it is true that the kind of signal being reported is consistent with that expected from simulations, which is reassuring at least. These debates should be settled with future, follow-up work. The ultimate arbiter should be the James Webb Space Telescope, which will hopefully be launched in 2013 and which will be powerful enough to identify individual POP III stars, rather than their imprints on the CIB.
