0221-Statistical properties of polarized radio sources at high frequency... arxiv:1003.5846

PAPER

Statistical properties of polarized radio sources at high frequency and their impact on CMB polarization measurements arxiv:1003.5846

Introduction

• Synchrotron radiation is the primary emission mechanism for sources detected in the radio band. Hence, radio sources are likely to be linearly polarized with a fractional polarization of typically a few percent, depending on the level of order of the magnetic field in the emission producing regions.

Conclusion

• the median level of fractional polarization is around 2 − 2.5 per cent and there is no evidence for any dependence on frequency in the range 8.4 to 43 GHz;

• if one is to detect r ∼ 0.01 some source mitigation strategy will be required at all frequencies.

• There is one final point we should make. Many of the sources, particularly those with high flux densities, are expected to be significantly variable. This will manifest itself not only in variable total and polarized flux densities, but also in the polarization position angle. This could make it difficult to subtract the effects of source confusion from the CMB polarization signal. If the observations of a particular CMB field are performed over a timescale which is shorter than the timescale of variability, this will require the high resolution observations performed to assist source subtraction to take place contemporaneously in order to make an accurate subtraction. Conversely, if the timescale for observations is longer than the variability, for example, if observations on a particular region are built up a series of short integrations over many days, then the variability of the source polarization could easily average out to a significantly lower observed polarization when the individual integrations are stacked. This effect will make accurate source subtraction difficult since it is likely to be impractical to monitor the level of variability for a substantial number of sources.

Source subtraction for the extended Very Small Array and 33-GHz source count estimates arxiv:0412605

Introduction

• Simple extrapolation of flux densities from low-frequency data can cause worse source contamination than if no subtraction were attempted at all, due to the population of sources with rising spectra

• ‘flat’-spectrum (flux-density spectral index α ≈ 0, S ∝ ν−α) sources will increasingly dominate bright flux density-selected samples at higher frequencies. Flat-spectrum sources are compact objects and often show flux density variations which means that accurate source subtraction at higher frequencies requires monitoring of source flux densities simultaneously with CMB observations.

• in order to avoid confusion by the CMB, surveying for sources should be performed at high resolution where the CMB contribution is negligible.