Estimates of unresolved point sources contribution to WMAP 5

TL;DR

This paper estimates the contribution of unresolved point sources to the WMAP temperature power spectrum using radio survey data and stochastic extrapolation, providing an alternative to previous models and highlighting the importance of external datasets.

Contribution

It introduces a stochastic extrapolation method of radio sources from NVSS data to estimate unresolved point source contamination in WMAP data, improving understanding of their impact.

Findings

Predicted source counts match WMAP results for flux > 1 Jy.

Sky maps correlate with WMAP observations for sources > 0.2 Jy.

Steeper frequency dependence found for sources below WMAP threshold.

Abstract

We present an alternative estimate of the unresolved point source contribution to the WMAP temperature power spectrum based on current knowledge of sources from radio surveys in the 1.4-90 GHz range. We implement a stochastic extrapolation of radio point sources in the NRAO-VLA Sky Survey (NVSS) catalog, from the original 1.4 GHz to the ~ 100 GHz frequency range relevant for CMB experiments. With a bootstrap approach, we generate an ensemble of realizations that provides the probability distribution for the flux of each NVSS source at the final frequency. The predicted source counts agree with WMAP results for S > 1 Jy and the corresponding sky maps correlate with WMAP observed maps in Q-, V- and W- bands, for sources with flux S > 0.2 Jy. The low-frequency radio surveys found a steeper frequency dependence for sources just below the WMAP nominal threshold than the one estimated by the WMAP team. This feature is present in our simulations and translates into a shift of 0.3-0.4 \sigma in the estimated value of the tilt of the power spectrum of scalar perturbation, n_s, as well as \omega_c. This approach demonstrates the use of external point sources datasets for CMB data analysis.