Perhaps the most astonishing and revolutionary discovery in cosmology was Edwin Hubble's observation that galaxies are moving away from us. It provides the underpinning of the big bang picture of creation in which the universe is expanding, and has been for 13.7 billion years. But astronomers in the last century were quick to point out to Hubble, and to the theoreticians like Einstein and Lemaitre who modeled his data, that the observations really only find that galaxies appear red. While relativity does predict that galaxies in an expanding universe will appear red, other causes of redness might be at work -- for example, a radical idea called "tired light" in which light in a static universe just grows redder as it travels over cosmic distances towards us.
For over sixty years scientists have tried to determine whether tired light, or perhaps some other effect, might be responsible for the redness of galaxies rather than expansion. One method they used was to watch supernovae. If motion of the object (that is, expansion) is responsible for red galaxies, then these fast moving objects will manifest other effects of their relativistic speeds. Not only the frequency of their light but also the frequency of all their phenomenon will appear to us to be "red," that is, to be happening more slowly. Supernovae, for example, will appear to glow for longer times in galaxies with larger redshifts. And indeed, all of the early studies found that supernovae behaved consistently with this notion, and the idea of tired light gradually lost favor.
Supernova measurements are difficult, however, and subject to numerous uncertainties. For example, the brightest supernovae -- those seen at the farthest distances -- might naturally glow for longer times as a consequence of their extreme luminosities. Indeed, researchers have found that this and other effects do influence supernovae lifetimes, meaning that not all supernovae are exactly identical with each other. CfA astronomers
Stephane Blondin, Michael Wood-Vasey, Peter Challis, Bob Kirshner, and Chris Stubbs, along with 27 of their colleagues, have now completed a definitive study that addresses all of these issues and unambiguously excludes the tired-light hypothesis. They watched changes in the spectra of thirteen supernovae in very red galaxies as these supernovae faded away. The time-varying spectral details of the supernovae enabled the team to calibrate the intrinsic ages and luminosities of the supernovae, and provided an accurate measure of the age of the supernovae. The scientists find that red galaxies have supernovae whose timing does indeed appear to be slow, consistent with relativity and the rapid motion of the host galaxies.
Their result, which unambiguously rules out the tired-light hypothesis, is the most direct confirmation of the reality of relativistic expansion that has ever been made.