A survey of hundreds of thousands of active galaxies reveals an asymmetry far stronger than that expected from our own motion through the Universe.
To understand this, we need to look at the sky in every direction and count very distant objects. If the Universe is broadly uniform, their distribution should become regular on very large scales. Our own motion nevertheless creates a small apparent difference between one direction in the sky and the opposite direction. This variation is known as a dipole.
The galaxies M81 and M82 observed in infrared by WISE, against a background filled with stars.
Credit: NASA/JPL-Caltech/UCLA
The researchers used infrared observations from WISE, a space telescope that mapped the entire sky. They searched for active galactic nuclei. These very luminous regions surround giant black holes that absorb matter. After several selections designed to limit contaminating sources, their final sample contained 636,966 candidates.
Our motion has already been measured using the cosmic microwave background, light emitted when the Universe was very young. Its dipole corresponds to a speed of about 370 km/s relative to this radiation. The same motion should produce a slight asymmetry in the number of galaxies observed depending on the direction being viewed.
Yet the signal measured with the galactic nuclei is about 2.5 times stronger than this prediction. The discrepancy reaches 5.3 sigma, a statistical measure indicating that it is difficult to attribute it to a simple random fluctuation. The direction of the observed dipole also differs from that obtained from the cosmic background.
Sky maps of radio sources and WISE quasars showing a dipole anisotropy in their distribution.
Credit: Secrest et al. / University of Oxford
The team notably improved the way objects were selected using their infrared light. This matters because selection criteria can themselves slightly alter the measured asymmetry. The researchers therefore adjusted their calculation to account for this effect.
Several explanations are possible. Irregularities in the surveys can create a false signal. The distribution of relatively nearby matter may also play a role. Another possibility is that matter and the cosmic background do not define exactly the same rest frame.
The challenge, therefore, is to distinguish an observation-related effect from a genuine cosmological phenomenon. Other surveys, conducted with different instruments and methods, will be able to test the same asymmetry. Agreement between several independent catalogues would make it much harder to explain the result as a bias specific to WISE.