A spatially dependent correction of Gaia EDR3 parallax zero-point offset based on 0.3 million LAMOST DR8 giant stars

Abstract: We have studied the zero-point offset of Gaia early Data Release 3 (EDR3) parallaxes based on a sample of 0.3 million giant stars built from the LAMOST data with distance accuracy better than 8.5\%. The official parallax zero-point corrections largely reduce the global offset in the Gaia EDR3 parallaxes: the global parallax offsets are $-$27.9 $\mu$as and $-$26.5 $\mu$as (before correction) and $+$2.6 $\mu$as and $+$2.9 $\mu$as (after correction) for the five- and six-parameter solutions, respectively. The bias of the raw parallax measurements is significantly dependent on the $G$ magnitudes, spectral colors, and positions of stars. The official parallax zero-point corrections could reduce parallax bias patterns with $G$ magnitudes, while could not fully account the patterns in the spaces of the spectral colors and positions. In the current paper, a spatially dependent parallax zero-point correction model for Gaia EDR3 five-parameter solution in the LAMOST footprint is firstly provided with the advantage of huge number of stars in our sample.