Comparison of filters for the detection of point sources in Planck simulations

We study the detection of extragalactic point sources in 2D flat simulations for all the frequencies of the forthcoming ESA’s Planck mission. In this work, we have used the most recent available templates of the microwave sky: as for the diffuse Galactic components and the Sunyaev‐Zel’dovich clusters we have used the ‘Plank Reference Sky Model’; as for the extragalactic point sources, our simulations ‐ which comprise all the source populations relevant in this frequency interval ‐ are based on up-to-date cosmological evolution models for sources. To consistently compare the capabilities of different filters for the compilation of ‐ hopefully ‐ the most complete blind catalogue of point sources, we have obtained three catalogues by filtering the simulated sky maps with: the matched filter (MF), the Mexican Hat Wavelet (MHW1) and the Mexican Hat Wavelet 2 (MHW2), the first two members of the Mexican Hat Wavelet Family. For the nine Planck frequencies we show the number of real and spurious detections and the percentage of spurious detections at different flux detection limits as well as the completeness level of the catalogues and the average errors in the estimation of the flux density of detected sources. Allowing a 5 per cent of spurious detections, we obtain the following number of detections by filtering with the MHW2 an area equivalent to half of the sky: 580 (30 GHz), 342 (44 GHz), 341 (70 GHz), 730 (100 GHz), 1130 (143 GHz), 1233 (217 GHz), 990 (353 GHz), 1025 (545 GHz) and 3183 (857 GHz). Our current results indicate that the MF and the MHW2 yield similar results, whereas the MHW1 performs worse in some cases and especially at very low fluxes. This is a relevant result, because we are able to obtain comparable results with the well-known MF and with this specific wavelet, the MHW2, which is much easier to implement and use.