Gaia Data Release 2
Documentation release 1.2

European Space Agency (ESA)
Gaia Data Processing and Analysis Consortium (DPAC)
5 June 2019
Executive summary

The second Gaia data release, Gaia DR2, encompasses astrometry, photometry, radial velocities, astrophysical parameters (stellar effective temperature, extinction, reddening, radius, and luminosity), and variability information plus astrometry and photometry for a sample of pre-selected bodies in the solar system. The basic number statistics of the contents of Gaia DR2 are as follows:

Data product or source type Number of sources
Total (excluding Solar system) 1 692 919 135
Five-parameter astrometry (position, parallax, proper motion) 1 331 909 727
Two-parameter astrometry (position only) 361 009 408
ICRF3 prototype sources (link to radio reference frame) 2 820
Gaia-CRF2 extra-galactic sources (optical reference frame) 556 869
G-band (330–1050 nm) 1 692 919 135
GBP-band (330–680 nm) 1 381 964 755
GRP-band (630–1050 nm) 1 383 551 713
Median radial velocity over 22 months 7 224 631
Classified as variable 550 737
Variable type estimated 363 969
Detailed characterisation of light curve 390 529
Effective temperature Teff 161 497 595
Extinction AG 87 733 672
Colour excess E(GBP-GRP) 87 733 672
Radius 76 956 778
Luminosity 76 956 778
Solar system object epoch astrometry and photonetry 14 099

The basic quality statistics of the contents of Gaia DR2 are as follows (where the astrometric uncertainties as well as the Gaia-CRF2 alignment and rotation limits refer to epoch J2015.5 TCB):

Data product or source type Typical uncertainty
Five-parameter astrometry (position & parallax) 0.020.04 mas at G<15
0.1 mas at G=17
0.7 mas at G=20
2 mas at G=21
Five-parameter astrometry (proper motion) 0.07 mas yr-1 at G<15
0.2  mas yr-1 at G=17
1.2 mas yr-1 at G=20
3 mas yr-1 at G=21
Two-parameter astrometry (position only) 14 mas
Systematic astrometric errors (sky averaged) <0.1 mas
Gaia-CRF2 alignment with ICRF 0.02 mas at G=19
Gaia-CRF2 rotation with respect to ICRF <0.02 mas yr-1 at G=19
Gaia-CRF2 alignment with ICRF 0.3 mas at G<12
Gaia-CRF2 rotation with respect to ICRF <0.15 mas yr-1 at G<12
Mean G-band photometry 0.3 mmag at G<13
2 mmag at G=17
10 mmag at G=20
Mean GBP- and GRP-band photometry 2 mmag at G<13
10 mmag at G=17
200 mmag at G=20
Median radial velocity over 22 months 0.3 km s-1 at GRVS<8
0.6 km s-1 at GRVS=10
1.8 km s-1 at GRVS=11.75
Systematic radial velocity errors <0.1 km s-1 at GRVS<9
0.5 km s-1 at GRVS=11.75
Effective temperature Teff 324 K
Extinction AG 0.46 mag
Colour excess E(GBP-GRP) 0.23 mag
Radius 10%
Luminosity 15%
Solar system object epoch astrometry 1 mas (in scan direction)

The above uncertainties on the photometry refer to the mean magnitudes listed in the main Gaia DR2 catalogue. The Gaia DR2 parallaxes show evidence for a global zeropoint, in the sense Gaia - ’true’, of about -0.03 mas, which has not been ’corrected’ in the data (for details, see Lindegren et al. 2018; Arenou et al. 2018).

The data collected during the first 22 months of the nominal, five-year mission have been processed by the Gaia Data Processing and Analysis Consortium (DPAC), resulting into this second data release. A summary of the release properties is provided in Gaia Collaboration et al. (2018b). The overall scientific validation of the data is described in Arenou et al. (2018). Background information on the mission and the spacecraft can be found in Gaia Collaboration et al. (2016), with a more detailed presentation of the Radial Velocity Spectrometer (RVS) in Cropper et al. (2018). In addition, Gaia DR2 is accompanied by various, dedicated papers that describe the processing and validation of the various data products: Lindegren et al. (2018) for the Gaia DR2 astrometry, Riello et al. (2018) and Evans et al. (2018) for the Gaia DR2 photometry, Sartoretti et al. (2018), Soubiran et al. (2018), and Katz et al. (2018) for the Gaia DR2 spectroscopy (radial velocities), Holl et al. (2018) for the Gaia DR2 variability, Andrae et al. (2018) for the Gaia DR2 astrophysical parameters, Gaia Collaboration et al. (2018f) for the Solar-system objects, and Mignard et al. (2018) for the celestial reference frame. Four more papers present a glimpse of the scientific richness of the data in the areas of the Hertzsprung-Russell diagram (Gaia Collaboration et al. 2018a), the mapping of the kinematics and large-scale structure of the Milky Way (Gaia Collaboration et al. 2018e), parallaxes and proper motions of Milky Way satellite galaxies (Gaia Collaboration et al. 2018d), and variable stars in the colour-magnitude diagram (Gaia Collaboration et al. 2018c). In addition to the set of references mentioned above, this documentation provides a detailed, complete overview of the processing and validation of the Gaia DR2 data.

Gaia data, from both Gaia DR1 and Gaia DR2, can be retrieved from the Gaia archive, which is accessible from The archive also provides various tutorials on data access and data queries plus an integrated data model (i.e., description of the various fields in the data tables). In addition, Luri et al. (2018) provide concrete advice on how to deal with Gaia astrometry, with recommendations on how best to estimate distances from parallaxes. The Gaia archive features an enhanced visualisation service, described in Moitinho et al. (2018), which can be used for quick initial explorations of the entire Gaia DR2 data set. Pre-computed cross matches between Gaia DR2 and a selected set of large surveys is provided, with details described in Marrese et al. (2019).

Gaia DR2 represents a major advance with respect to Gaia DR1 in terms of survey completeness, precision and accuracy, and the richness of the published data. Nevertheless, Gaia DR2 is still an early release based on a limited amount of input data, simplifications in the data processing, and imperfect calibrations. Many limitations hence exist (as described in Gaia Collaboration et al. 2018b) and we summarise here the most important ones that the user of Gaia DR2 should be aware of:

  • The validation of Gaia DR2 was done in various stages and led to the decision to filter out parts of the available data processing results before publication. The applied filters are summarised in Gaia Collaboration et al. (2018b) and the details can be found in the data processing papers listed above, and in this documentation (Sections 4.4.3, 6.2.1, and 9.2.3, and Chapter 10).

  • The survey represented by Gaia DR2 is essentially complete between G=12 and G=17 mag. At the bright end, the completeness has improved compared to Gaia DR1. At G<7 mag, however, there are still many sources missing from the catalogue, primarily due to the difficulties of treating saturated CCD images. Fainter than G=17 mag, the completeness is affected by a combination of data processing limitations in crowded fields and the filtering applied before publication. Filtering based on the number of observations of a given source will lead to imprints of the scanning law in the celestial distribution of sources which is noticeable at G>17 mag. The various subsets of Gaia DR2, such as the radial velocity or variability surveys, are all incomplete to varying degrees with respect to the overall catalogue. More details on the survey completeness can be found in Gaia Collaboration et al. (2018b), Arenou et al. (2018), the papers listed above, and in Chapter 10. No attempt was made at deriving a selection function for Gaia DR2.

  • During the data processing for this release, including the astrometric and radial velocity processing, all sources were treated as single stars. This means that, for binary and multiple stellar systems, the astrometry and median radial velocity will be less accurate. Sources that are not single stars are not marked as such in the catalogue. The auxiliary information in Gaia DR2 does allow the identification of, for example, extragalactic sources but this should be done with care and only after verification on known samples.

  • The systematic errors in the parallaxes are estimated to be below the 0.1 mas level (Lindegren et al. 2018) but the following systematics remain. There is an overall parallax zeropoint which, from an examination of QSO parallaxes, is estimated to be around -0.03 mas (in the sense of the Gaia DR2 parallaxes being too small). The estimated parallax zeropoint depends on the sample of sources examined (Arenou et al. 2018) and the value above should not be used to ’correct’ the catalogue parallax values. In addition to a global zeropoint, there are also regional systematic errors as well as source-to-source correlations in the errors (which also affect the other astrometric parameters). The astrometric uncertainties are known to be underestimated and the bright star (G13 mag) astrometry is still limited by calibration uncertainties. More information is provided in Lindegren et al. (2018), Arenou et al. (2018), and Chapter 10.

  • Systematic errors also exist in the photometry (Evans et al. 2018) and in the radial velocities (Katz et al. 2018), where we note in particular a positive bias, with respect to other radial velocity surveys, which increases with source magnitude up to around 500  m s-1 at GRVS12 mag (see Section 6.5.2). The astrophysical parameter estimates in Gaia DR2 suffer from systematics related to the strong assumptions necessary to derive effective temperature and extinction from the Gaia broad band photometry and parallaxes only as well as limitations in the model grids that were used to train the machine learning algorithms employed in the data processing (Andrae et al. 2018).

  • The photometric colour information in Gaia DR2 suffers from systematics at the faint (G>19 mag) end of the survey, in crowded regions, and in the vicinity of bright stars. In these cases, the removal of the astronomical and stray light background contributions from the integrated fluxes in the BP and RP prism photometers was inaccurate, leading to overestimates of the fluxes in GBP and GRP. The result is that source colours for faint stars, in crowded regions, and in the surroundings of bright stars are unreliable. One should treat colour-magnitude diagrams constructed for these cases with care. The Gaia DR2 catalogue provides a field, phot_bp_rp_excess_factor, which can be used to judge the extent to which the photometry of a given source is compromised. More detailed information and guidance on cleaning samples from unreliable photometry is provided in Evans et al. (2018), Arenou et al. (2018), Gaia Collaboration et al. (2018a), and Section 5.5.2 and Chapter 10 of this documentation.

  • Radial velocities are only reported for stars for which the template spectrum used in the cross-correlation method has an effective temperature in the range 35506900 K. Radial velocities with absolute values above 550  km s-1 should be treated with care. See Katz et al. (2018) and Section 6.5.2 for more details.

  • The astrophysical parameters Teff, AG, E(GBP-GRP), radius, and luminosity should be used with care. It is strongly recommended to follow the guidelines for using these data provided in Andrae et al. (2018).

  • For closely aligned sources (separated by 0.20.3 arcsec), which are only occasionally resolved in the Gaia observations, confusion in the observation-to-source matching can lead to spurious parallax values which are either very large or have a negative value very far away from zero in terms of the formal parallax uncertainty quoted in the catalogue. These sources tend to be faint and located in crowded regions and are also associated with unreliable (large) proper motions (Gaia Collaboration et al. 2018b). Guidance on how to clean samples from spurious parallax values is provided in Lindegren et al. (2018).

  • Gaia DR2 and Gaia DR1 should be treated as two completely independent catalogues. This is explained in detail in Gaia Collaboration et al. (2018b). In particular the photometric systems of the two catalogues are different (see also Evans et al. 2018) and the source list has changed substantially in the sense that many sources (some 80–90% at G<16 mag) have changed source identifier between Gaia DR1 and Gaia DR2. This means that one cannot look up sources from Gaia DR1 in Gaia DR2 on the basis of the source identifier. A table to identify the same physical source in the two data releases is provided in the Gaia DR2 archive. For samples of sources one should not use a previously calculated cross-match with Gaia DR1 but calculate a new cross-match with Gaia DR2. For several large surveys, pre-computed cross-matches are already provided in the Gaia DR2 archive (see Marrese et al. 2019, and Section 14.5).

Summary of miscellaneous links:

List of Figures:
List of Tables: