A new technique for reliably identifying point sources in millimeter/sub-millimeter wavelength maps is presented. This method accounts for the frequency dependence of noise in the Fourier domain as well as non-uniformities in the coverage of a field.
This optimal filter is an improvement over commonly-used matched filters that ignore coverage gradients. Treating noise variations in the Fourier domain as well as map space is traditionally viewed as a computationally intensive problem. We show that the penalty incurred in terms of computing time is quite small due to casting many of the calculations in terms of FFTs and exploiting the absence of sharp features in the noise spectra of observations. Practical aspects of implementing the optimal filter are presented in the context of data from the AzTEC bolometer camera. The advantages of using the new filter over the standard matched filter are also addressed in terms of a typical AzTEC map.
We present a 1.1 mm wavelength imaging survey covering 0.3 sq. deg. in the COSMOS field. These data, obtained with the AzTEC continuum camera on the James Clerk Maxwell Telescope (JCMT), were centred on a prominent large-scale structure over-density
which includes a rich X-ray cluster at z = 0.73. A total of 50 millimetre galaxy candidates, with a significance ranging from 3.5-8.5 sigma, are extracted from the central 0.15 sq. deg. area which has a uniform sensitivity of 1.3 mJy/beam. Sixteen sources are detected with S/N > 4.5, where the expected false-detection rate is zero, of which a surprisingly large number (9) have intrinsic (de-boosted) fluxes > 5 mJy at 1.1 mm. Assuming the emission is dominated by radiation from dust, heated by a massive population of young, optically-obscured stars, then these bright AzTEC sources have FIR luminosities > 6 x 10^12 L(sun) and star formation-rates > 1100 M(sun)/yr. Two of these nine bright AzTEC sources are found towards the extreme peripheral region of the X-ray cluster, whilst the remainder are distributed across the larger-scale over-density. We describe the AzTEC data reduction pipeline, the source-extraction algorithm, and the characterisation of the source catalogue, including the completeness, flux de-boosting correction, false-detection rate and the source positional uncertainty, through an extensive set of Monte-Carlo simulations. We conclude with a preliminary comparison, via a stacked analysis, of the overlapping MIPS 24 micron data and radio data with this AzTEC map of the COSMOS field.
