Tuesday 02 May 2017: The Effect of Unresolved Contaminant Stars on the Cross-Matching of Photometric Catalogues
Tom Wilson - University of Exeter
4th Floor Interaction Area 11:15-11:45
In this talk I will discuss the effects of crowding in photometric surveys on both the astrometric and photometric properties measured. The blending of stellar objects has implications for the distribution of cross-match separations. Blended objects exhibit large, non-Gaussian separation tails, for catalogue densities of stars per point spread function area as low as 0.5%. These perturbations can be on the order of several arcseconds, in cases of large point spread functions. A more general probability-based matching scheme has the flexibility to model non-Gaussian distributions, combining the rejection of false matches through magnitude comparison, which I will briefly discuss. This allows for the recovery of objects of high astrometric perturbation and the avoidance of matches below the sensitivity limit of one survey. Finally, I will also consider the potential for unphysical excesses in the stellar fluxes through photometric contamination, using WISE, vital for the mid-IR disk detections, as an example. At low Galactic latitudes, stars as bright as W1=13 will, on average, have at least one additional source of flux in their point spread function. This leads to approximately 50% of Gaia-WISE cross-matches having separations larger than their 3-sigma uncertainty matching radius, assuming traditional Gaussian statistics. This means that half of the potential matches would be rejected by a Gaussian-based probability-matching scheme, with some objects perturbed even beyond a 3 arcsecond nearest neighbour matching radius. These lost matches suffer, on average, 20% flux contamination.