Statistical properties of extragalactic sources in the New Extragalactic WMAP Point Source (NEWPS) catalogue

TL;DR

This paper analyzes the statistical properties of extragalactic sources in the NEWPS catalogue, including spectral indices, counts, and redshifts, providing insights into their spectral behavior and consistency with cosmological models.

Contribution

It offers a comprehensive statistical analysis of extragalactic sources in the NEWPS catalogue, including spectral, count, and redshift distributions, with calibration and classification methods.

Findings

76% of sources are flat-spectrum between 5 and 23 GHz

Spectral steepening observed at higher frequencies

Source counts agree with cosmological evolution models

Abstract

We present results on spectral index distributions, number counts, redshift distribution and other general statistical properties of extragalactic point sources in the NEWPS5 sample L\'opez-Caniego et al. (2007). The flux calibrations at all the WMAP channels have been reassessed both by comparison with ground based observations and through estimates of the effective beam areas. The two methods yield consistent statistical correction factors. A search of the NED has yielded optical identifications for 89% of sources in the complete sub-sample of 252 sources with S/N>5 and S>1.1 Jy at 23 GHz; 5 sources turned out to be Galactic and were removed. The NED also yielded redshifts for 92% of the extragalactic sources at |b|>10deg. Their distribution was compared with model predictions; the agreement is generally good but a possible discrepancy is noted. Using the 5 GHz fluxes from the GB6 or PMN surveys, we find that 76% of the 191 extragalactic sources with S_23GHz>1.3,Jy can be classified as flat-spectrum sources between 5 and 23 GHz. A spectral steepening is observed at higher frequencies: only 59% of our sources are still flat-spectrum sources between 23 and 61 GHz and the average spectral indexes steepen from <alpha_5^23>= 0.01\pm 0.03 to <alpha_41^61>= 0.37\pm 0.03. We think, however, that the difference may be due to a selection effect. The source number counts have a close to Euclidean slope and are in good agreement with the predictions of the cosmological evolution model by De Zotti et al. (2005). The observed spectral index distributions were exploited to get model-independent extrapolations of counts to higher frequencies. The risks of such operations are discussed and reasons of discrepancies with other recent estimates are clarified.