[Paper Review] Gaia Early Data Release 3 -- Catalogue validation

C. Fabricius, X. Luri+34 more|Dipòsit Digital de la Universitat de Barcelona (Universitat de Barcelona)|Dec 11, 2020

Stellar, planetary, and galactic studies33 references22 citations

TL;DR

This paper validates the Gaia Early Data Release 3 (EDR3) catalogue, confirming significant improvements in astrometric and photometric precision, accuracy, and completeness over Gaia DR2. It identifies residual systematics—particularly in proper motions and photometry—provides correction recommendations, and offers data quality indicators to guide reliable scientific use of the catalogue.

ABSTRACT

The third Gaia data release is published in two stages. The early part, Gaia EDR3, gives very precise astrometric and photometric properties for nearly two billion sources together with seven million radial velocities from Gaia DR2. The full release, Gaia DR3, will add radial velocities, spectra, light curves, and astrophysical parameters for a large subset of the sources, as well as orbits for solar system objects. Before the publication of the catalogue, many different data items have undergone dedicated validation processes. The goal of this paper is to describe the validation results in terms of completeness, accuracy, and precision for the Gaia EDR3 data and to provide recommendations for the use of the catalogue data. The validation processes include a systematic analysis of the catalogue contents to detect anomalies, either individual errors or statistical properties, using statistical analysis and comparisons to the previous release as well as to external data and to models. Gaia EDR3 represents a major step forward, compared to Gaia DR2, in terms of precision, accuracy, and completeness for both astrometry and photometry. We provide recommendations for dealing with issues related to the parallax zero point, negative parallaxes, photometry for faint sources, and the quality indicators.

Motivation & Objective

• To assess the completeness, accuracy, and precision of the Gaia EDR3 catalogue across astrometry and photometry.
• To identify and characterize systematic errors and anomalies in the data, especially those persisting from Gaia DR2.
• To provide data quality indicators and correction recommendations for users to ensure reliable scientific interpretation.
• To validate the performance of photometric and astrometric parameters in crowded and faint-source regimes.
• To benchmark EDR3 against DR2 and external data, ensuring consistency and improved reliability for astronomical applications.

Proposed method

• Conducted systematic statistical analysis of catalogue contents to detect individual errors and statistical anomalies.
• Performed cross-comparisons between Gaia EDR3 and Gaia DR2, as well as with external astrophysical models and independent datasets.
• Utilized quality indicators such as ruwe and astrometric_gof_al to assess astrometric solution reliability, especially in crowded fields.
• Analyzed photometric performance using the phot_bp_rp_excess_factor to quantify flux consistency across BP and RP bands.
• Evaluated the impact of background subtraction and source blending on photometric accuracy, particularly for faint and red sources.
• Applied simulations and statistical modeling to quantify systematics in photometric measurements, especially for $G_{\rm BP}$ and $G_{\rm RP}$ at faint magnitudes.

Experimental results

Research questions

• RQ1 How has the astrometric precision and accuracy of Gaia EDR3 improved over Gaia DR2?
• RQ2 What residual systematics remain in the proper motions and astrometric parameters, and how do they vary with magnitude and source density?
• RQ3 To what extent are the $G_{\rm BP}$ and $G_{\rm RP}$ photometric measurements affected by background contamination and angular resolution limitations?
• RQ4 How do data quality indicators like ruwe and phot_bp_rp_excess_factor perform in identifying unreliable measurements, particularly in crowded regions?
• RQ5 What corrections are necessary for $G$ photometry in sources with default colours, and how effective are they?

Key findings

• Gaia EDR3 shows a significant reduction in the $G$ magnitude trend and weaker discontinuities at $G=10.87$ and $13$ mag compared to Gaia DR2.
• Proper motions in right ascension exhibit a small systematic offset of 0.01 mas yr⁻¹ for sources brighter than $G=13$ mag.
• The $G_{\rm BP}$ and $G_{\rm RP}$ photometry is less affected by background subtraction errors than in Gaia DR2, but completeness is reduced in dense fields.
• Photometry in $G_{\rm BP}$ fainter than 20.5 mag and $G_{\rm RP}$ fainter than 20 mag is biased toward brighter values due to systematics in flux measurement.
• The use of $G - G_{\rm RP}$ instead of $G_{\rm BP} - G_{\rm RP}$ is recommended for faint, red sources to avoid systematic biases.
• Systematic errors related to the scanning law pattern persist in astrometric parameters, though photometric systematics have improved significantly compared to Gaia DR2.