Systematics and accuracy of VLBI astrometry: A comparison with Gaia Data Release 2

¹ School of Astronomy and Space Science, Key Laboratory of Modern Astronomy and Astrophysics (Ministry of Education), Nanjing University, 163 Xianlin Avenue, 210023 Nanjing, PR China
e-mail: liuniu@smail.nju.edu.cn; zhuzi@nju.edu.cn
² SYRTE, Observatoire de Paris, Université PSL, CNRS, Sorbonne Université, LNE, Paris, France
e-mail: sebastien.lambert@obspm.fr

Received: 25 October 2019
Accepted: 13 December 2019

Abstract

Context. The third generation of the ICRF – ICRF3 – was published in 2018. This new fundamental catalog provides radio source positions measured independently at three bands: S/X, K, and X/Ka, representing three independent radio celestial frames which altogether constitute a multi-frequency ICRF.

Aims. We aim to investigate the overall properties of the ICRF3 with the help of the Gaia Data Release 2 (Gaia DR2). This could serve as an external check of the quality of the ICRF3.

Methods. The radio source positions of the ICRF3 catalog were compared with the Gaia DR2 positions of their optical counterparts at G <  18.7. Their properties were analyzed in terms of the dependency of the quoted error on the number of observations, on the declination, and the global difference, the latter revealed by means of expansions in the vector spherical harmonics.

Results. The ICRF3 S/X-band catalog shows a more smooth dependency on the number of observations than the ICRF1 and ICRF2, while the K and X/Ka-band yield a dependency discrepancy at the number of observations of ∼50. The rotation of all ICRF catalogs show consistent results, except for the X-component of the X/Ka-band which arises from the positional error in the non-defining sources. No significant glides were found between the ICRF3 S/X-band component and Gaia DR2. However, the K- and X/Ka-band frames show a dipolar deformation in Y-component of +50 μas and several quadrupolar terms of 50 μas in an absolute sense. A significant glide along Z-axis exceeding 200 μas in the X/Ka-band was also reported. These systematics in the ICRF catalog are shown to be less dependent on the limiting magnitude of the Gaia sample when the number of common sources is sufficient (>100).

Conclusions. The ICRF3 S/X-band catalog shows improved accuracy and systematics at the level of noise floor. But the zonal errors in the X/Ka-band should be noted, especially in the context of comparisons of multi-frequency positions for individual sources.