CMB Spectral Distortions Revisited

While uploading some bibliographic information for bureaucratic purposes yesterday I noticed that an old paper of mine had recently attracted a number of citations. The paper was written while I was a postdoctoral research fellow in the Astronomy Centre at the University of Sussex in 1990, but not published until 1991 by which time I had moved to Queen Mary College (as it was then called). The citation history of this article is actually quite interesting:

You can see that it was cited a bit immediately after publication, then endured a long spell from 1997 to 2012 in which nobody seemed interested in it, then experienced something of a revival. It currently has a total of about 49 citations, which doesn’t exactly make it a classic in a field which is extremely active, but it’s nice to see it hasn’t been forgotten entirely.

Here is the abstract of the paper:

As the abstract makes clear we wrote this paper in response to a measurement of the spectrum of the cosmic microwave background radiation by the FIRAS instrument on the satellite COBE that had demonstrated that it was extremely well fitted by a Planck spectrum, with little room for any deviation away from a perfect black-body shape. Here’s the measured curve from COBE and some other experiments at the time:

The accuracy of the fit allows one to place limits on any process happening in the early Universe that might produce a distortion of the spectrum. There are a number of things that could do this. Any energy released in the early Universe takes time to thermalise, i.e. for the radiation field and the matter to come into thermal equilibrium via Compton scattering, double Compton scattering and Bremsstrahlung. Imperfect thermalisation produces a spectrum which doesn’t quite match the Planck curve.

Two types of distortion are possible, both introduced in classic papers from 1969 and 1970 by Rashid Sunyaev and Ya. B. Zel’dovich. One type is called a y-distortion (which corresponds to photons being shifted from low frequency and the other is called a μ-distortion, which is described by inserting a chemical potential term to the usual Planck formula for the black-body spectrum. Observational limits on both forms of distortion are very tight : |y|<1.5 ×10^-5; |μ|<1.5 ×10^-5, which places stringent limits on any energy release, including that which would arise from the dissipation of primordial acoustic waves (which is what John and I concentrated on in the paper).

So why did interest in this get revived a few years ago? The answer to that is that advances in relevant technology have now made it possible to think about an experiment that can measure much smaller spectral distortions than has hitherto been possible. A proposal for an experiment, called PIXIE, which includes such a measurement, is described here. Although spectral distortions are only a secondary science goal for PIXIE, it could push down the upper limits quoted above by a factor of 1000 or so, at which level we should expect to see departures from the Planck curve within the standard model, which would be a very important test of basic cosmological theory.

That all depends on whether PIXIE – or something like it – goes ahead.

Follow @telescoper