CAOSP abstracts, Volume: 56, No.: 3, year: 2026

Author(s): Li, F.; Huang, G.; Yang, L.
Journal: Contributions of the Astronomical Observatory Skalnaté Pleso, vol. 56, no. 3, p. 35-53.
Date: 08/2026
Title: A Gaia DR3--LAMOST reassessment of high-velocity star candidates with geometric-distance-aware kinematics
Keyword(s): high-velocity stars, Galaxy: kinematics and dynamics, stars: distances, catalogues, methods: data analysis
Pages: 35 -- 53
DOI: https://doi.org/10.31577/caosp.2026.56.3.35
NASA ADS Abstract Service: 2026CoSka..56c..35L
Citation:
License: This work is licensed under a Creative Commons Attribution 4.0 International License

Abstract: We revisit three published high-velocity-star catalogues drawn from Gaia and LAMOST -- Li et al. (2021, 591 stars), Li et al. (2023, 88 stars) and Liao et al. (2024, 519 stars) -- and quantify how the published unbound classification of these candidates changes with the choice of distance estimator and with the choice of radial-velocity catalogue. After Gaia DR3 source_id resolution and de-duplication our master sample contains 1101 unique stars. 675 of them satisfy ruwe < 1.4, ϖ/σ_ϖ > 5 and carry a Gaia DR3 radial velocity (the Gaia-only clean sample). 356 additionally pass a 1 arcsec LAMOST DR9 LRS cross-match, atmospheric-parameter cuts and a 50 km s^-1 Gaia--LAMOST RV agreement test (the final-strict sample). For every star in this strict subset, the three-dimensional Galactocentric rest-frame speed v_grf and the local escape speed v_esc are computed with a 1000-draw Monte Carlo under the MWPotential2014 potential of Bovy (2015), once with inverse-parallax distances (1000/ϖ) and once with the geometric distances of Bailer-Jones et al. (2021). Switching from inverse parallax to Bailer-Jones distances on the same 356 stars reduces the count with P_unbound = P(v_grf > v_esc) > 0.5 from 48 to 3, and the count with P_unbound > 0.9 from 12 to 1; 45 stars are downgraded across the P_unbound = 0.5 threshold and none are gained. Substituting LAMOST radial velocities for Gaia radial velocities on the same subset changes the median v_grf by 0.00 km s^-1 (p₉₀|Δv_grf| = 5.7 km s^-1) and flips the P_unbound = 0.5 classification of one star out of 356. The unbound classification of these published candidates is therefore set by the distance estimator, not by the choice of radial-velocity catalogue. The final-strict sample, a Top-15 main candidate table, and a Top-30 machine-readable supplementary CSV are released with this paper; the three highest-confidence candidates are metal-poor giants that warrant targeted spectroscopic follow-up.