Speaker
Description
The Gaia DR3 astrometric selection contains about 1.9 million quasar candidates whose astrometry is consistent with an extragalactic origin. In a first study, we examined how the normalised parallaxes and proper motions of this sample are distributed across more than 800 subsamples defined by Gaia's own solution quality indicators and source parameters, such as ruwe, the type of astrometric solution, magnitude and morphology. The width and the skewness of these distributions serve as two complementary measures of how closely a subsample follows the behaviour expected of quasars. The analysis reveals a better subset, mainly five-parameter solutions of point-like sources with good fit statistics and without sources selected only through external catalogues (not confirmed by the Gaia DR3 classifiers), and the remaining, worse subset. Using the skewness and the corresponding distributions for confusion sources, we also attempt an indicative estimate of the residual stellar contamination in the astrometric selection. In a second study, we use our findings to extend the analysis from subsets to individual sources. Each source is assigned scores based on the behaviour, in terms of width and skewness, of the subsets it belongs to, and the sample is ordered from best to worst and divided into quintiles. We then follow how parameters such as fit statistics, solution type, magnitude, and colour change along this ranking and clarify which parameters matter most for building the score. The final aim is a single score that reflects both measures. Our ranking of the DR3 quasars allows one to select subsets of these quasars, trading off sample size against astrometric quality.