The Odd Spatial Distribution of Radio Galaxies on Cosmic Scales

Radio-loud Active Galactic Nuclei (AGNs) are among the brightest extragalactic sources and can be observed to very large distances. They already existed when the universe was only half its current age and thus should bear witness of the onset of the accelerated expansion of the universe. Radiation from the Cosmic Microwave Background (CMB) reaches us from even earlier times and has to traverse the space around the AGNs. As the accelerated expansion causes the gravitational potential wells on large scales - where the AGNs are sitting — to become shallower, anisotropies should result in the CMB radiation. These anisotropies should be spatially correlated with the AGN distribution. Contrary to expectations, however, scientists at the Max Planck Institute for Astrophysics find no similarity between anisotropies in the CMB and the angular patterns of a radio source catalogue. Even more surprising, they find a clear mismatch between the theoretically expected and the observed distribution of radio sources on large scales.

Fig. 1: Simulated CMB map (top) and simulated projected density map mimicking the spatial distribution for the NVSS AGN catalogue (bottom). As expected from theory, both maps are similar particularly on large scales. In these mock maps only fluctuations on angular scales larger than ~12 degrees are shown, which makes these similarities visible even to the naked eye. Grey areas are discarded in the analysis of real data (see Fig. 2).

Fig. 2: Maps of real WMAP 7th year (V-band) CMB data (top) and the projected density of real AGNs in the NVSS catalogue (bottom). Grey areas are discarded in the analysis due to either a lack of data or high contamination by other astrophysical sources. There does not seem to be any similarity (contrary to what is shown in Fig.1), which is confirmed by a statistical analysis.

Fig. 3: Amount of angular fluctuations in the NVSS AGN distribution for different angular multipoles (l). At low multipoles (corresponding to large angular scales) there seems to be a clear excess when comparing to theoretical predictions (black solid line). Black, red and green symbols correspond to AGN at different flux thresholds: 2.5, 30 and 60 mJy, respectively. At high multipoles (corresponding to small angular scales), the excess of red and green symbols (corresponding to 30 and 60 mJy) is well understood: This effect arises because of the relatively small number of AGNs above those flux cuts.

Radio-loud AGNs are galaxies that host a super-massive black hole in their centre that is accreting material from its surroundings. These AGNs have been found to very large distances, corresponding to times when our universe was less than half its current age. Other observations indicate that around this time, the universe becomes dominated by a new component different from matter and radiation, called "Dark Energy", which counteracts the attractive forces of gravity and hence accelerates the expansion rate of the Universe.

This Dark Energy is subject of intensive research. Major on-going and upcoming optical surveys attempt to characterize its nature by investigating how it modifies the lensing pattern of galaxies or their clustering, or by directly measuring the expansion rate at different cosmological epochs. One clear cut prediction for Dark Energy is that it should modify the growth of gravitational potential wells on the largest scales, if it indeed accelerates the expansion rate of the universe. In a Universe with a flat geometry such as ours, in the absence of Dark Energy gravitational potential wells should remain constant: their growth due to gravitational enhancement of overdensities is perfectly cancelled by the expansion rate of space. However, if Dark Energy becomes dominant and accelerates the universal expansion rate, then these gravitational potential wells should become shallower, at least on the largest scales.

This shrinking of the potential wells should have an observable effect on the intensity anisotropies of the Cosmic Microwave Background (CMB). This radiation was emitted at very early epochs of our universe, roughly 380,000 years after the Big Bang, and has crossed the whole visible universe when it reaches us. In particular, the CMB photons must have crossed the evolving gravitational potential wells. If these have become shallower during the time required for CMB photons to cross those wells, then these photons should have gained some energy, since they leave a potential well that is shallower at exit than it was at entrance. This gravitational "blue-shift" of the CMB is known as the Integrated Sachs-Wolfe effect (hereafter ISW).

The number of gravitational potential wells varies along different directions on the sky, which means that the ISW effect will introduce angular anisotropies on the intensity pattern of the CMB. Now, if galaxies and AGNs preferentially form in those same potential wells, then there should be some correlation between the angular pattern of the CMB and the AGN angular distribution. Figure 1 shows this similarity in two simulated maps of the CMB radiation and the angular density fluctuations of radio galaxies after filtering out small scale anisotropies. On large angular scales (corresponding to large gravitational potential wells) both maps show some resemblance, which can be measured with high significance by applying a statistical analysis.

Applying the same statistics to real observational data, however, shows no significant similarity between these two maps. Figure 2 shows real CMB data (as measured by the WMAP satellite) and the angular distribution of objects in the NRAO VLA Sky Survey (NVSS) catalogue, which contains some 1.6 million extragalactic radio sources, of which more than 99% should be AGNs.

What does this mismatch mean? The CMB, on the one hand, has been compared to theoretical expectations and has proved to be in good agreement overall. On the other hand, the fluctuations of NVSS sources on large scales show a clear excess when compared to predictions. The fact that this excess is present for different flux thresholds (i.e. also brightest, clearly detected sources, show this abnormal behaviour) makes any observational systematic unlikely - but not impossible.

Is this excess of large scale anisotropy of radio AGN comparable to that recently claimed in the Sloan catalogue of Luminous Red Galaxies? Is this excess a signature of intrinsic non-Gaussianity in the matter fluctuation field in our universe? Do the AGNs in NVSS really sample the large scale gravitational wells? Can we derive conclusions on the mysterious Dark Energy from this result? These are all exciting, open questions, which are the subject of current investigations.


Carlos Hernandez-Monteagudo


Further information:

Carlos Hernandez-Monteagudo, "Revisiting the WMAP-NVSS angular cross correlation. A skeptic's view", Astronomy & Astrophysics, 520, 101 (2010), linkPfeilExtern.gifarxiv:0909.4294