FAQ on Gaia Data Access


When using the Gaia Archive, you might not be aware of some essential information that could help you to use the Gaia Archive more effectively. Also, you might encounter an issue that is easily solvable by yourself. Here you can find this type of information. When things are still unclear, we welcome you to raise your question at the Gaia Helpdesk.


Upcoming upgrades to the Gaia Archive

Our Gaia Archive team is continuously working towards making the Gaia Archive more efficient and easy to use. New releases of the Gaia Archive come out every now and then. Service interruptions are announced beforehand in the GUI (one-time pop-ups). On the Gaia Archive "New Releases" page you can find information on the current and previous releases.


First aid solution to Access issues

Planned maintenance to the Gaia Archive is announced through Gaia Cosmos and/or through a pop-up warning at the Gaia Archive. If you cannot access the Gaia Archive, and the unavailability of the Gaia Archive is not announced there, please contact us through the Gaia Helpdesk.

The first aid solution to most issues with the Gaia Archive GUI (from seemingly disappeared tables or strange behaviour after a query or problems signing in or out), is to close your browser, clear your cache and then retry. In most cases, this will solve things and you will be able to continue working in no time.

A second solution might be to switch browsers. Some browsers show more hickups than others in combination with the use of the Gaia Archive GUI.


First aid solution to exceeded file quota

Assumed is that you entered the Gaia Archive with a registered account. A registered account has the benefits of a dedicated user space with more space and larger quota. If you get an error message that you have exceeded your file quota, but you did not according to the information in your account, first check if you have jobs that failed to complete. You could see whether deleting the jobs that failed solve the issue. Even though they seem to take up no space (0 bytes), sometimes they do and deleting them will free up space.

Another quick test to try and see if you can solve the exceeded file quota yourself, is to clear your cache (so close your browser and start a new private window) and then check again.


First aid solution to questions

Perform a search on the internet with your question. You probably find that the answer to your question is already around on one of the Gaia information pages. Maybe you can find the answer as one of the FAQ on Gaia data access below, or there might be an answer available already from the Gaia FAQ (answering questions on the Gaia Mission and Data Releases).

Please also have a look at the Gaia Archive help pages which contains lots of helpful information and tutorials or perform a search through the Gaia Data Release Documentation and the Gaia Data Release papers (Gaia DR1, Gaia DR2). The question might be hidden in the information on some other Gaia Cosmos page or Gaia pages of our DPAC partners. When not sure what information is around for a certain data release, check out the overview pages for each release: Gaia Data Release 1 and Gaia Data Release 2.


Gaia helpdesk

For all questions on the Gaia Archive, use of the Gaia Archive, content of the Gaia Archive and more, feel free to contact the Gaia Helpdesk.


FAQ on Gaia Data Access


Where can I find the column descriptions and units of the data fields?

The Gaia Archive data model (column description, units, etc.) along with the extensive documentation of the data and its processing can be found here, with the opportunity to download the full data release documentation.


Why are there duplicate Hipparcos matches in table public.igsl_source_catalog_ids?

This is the unavoidable consequence of using the Initial Gaia Source List (IGSL) for bootstrapping the cross-match procedures for Gaia Data Release 1. The IGSL is a collection of multiple, incomplete pre-Gaia catalogues, known to contain spurious and duplicate sources. To avoid dropping Hipparcos stars in the IGSL, those entries not matched were added as a "fake" Tycho-2 star with Tycho-2 identifier idTYCHO = 9999999000000+idHIP. Because of a bug in the matching procedures, approximately 12000 Hipparcos stars have been entered twice. These can be identified as objects with auxHIP = 1 and idTYCHO > 9999999000000 and should not be used. 


What are the celestial coordinates of my HEALPix identifier?

This table provides a link between the HEALPix identifier (level 6, so Nside = 2^6 = 64) and equatorial, Galactic, and ecliptic coordinates. The text file has four header lines and 49152 lines, each with seven columns separated by commas:

  1. HEALPix identifier [nest scheme, Gaia default]
  2. Right ascension [deg]
  3. Declination [deg]
  4. Galactic Longitude [deg]
  5. Galactic Latitude [deg]
  6. Ecliptic Longitude [deg]
  7. Ecliptic Latitude [deg]
  8. ​​​​​​​HEALPix identifier [ring scheme]

The conversion from Gaia source_id to HEALPix number with nest pixel ordering is described here. The built-in Archive function GAIA_HEALPIX_INDEX(order, source_id) can be used to do the transformation within ADQL queries, more details at the Gaia Archive help => ADQL syntax => 3. Additional functions available. Read here about ring and nested HEALPix ordering.


Why does my query time out after 90 minutes? / Why is my query limited to return 3 million rows?

There are limits on how much data can be retrieved and how long a query can last. They are different for anonymous and authenticated users.

Anonymous access (as documented here):

  • ADQL synchronous query time-out: 30 s
  • ADQL asynchronous query time-out: 90 min
  • ADQL query maximum number of rows: 3,000,000
  • Datalink (e.g. light curves) maximum number of sources: 1000

Authenticated users:

  • ADQL synchronous query time-out: 60 s
  • ADQL asynchronous query time-out: 120 min
  • ADQL asynchronous query maximum number of rows: unlimited
  • Jobs (ADQL query output): 20 GB
  • User tables: 1 GB
  • Datalink (e.g. light curves) maximum number of sources: 1000

Anybody can self-register through the "SIGN IN" button on the top-right corner of the Gaia Archive, after which the reported user quota apply. Upon demonstrated need, users can request (temporary) changes to their quota by sending a motivated request to the Gaia Helpdesk.


I can not find a well known bright star in the Gaia Archive. Why?

These stars might be too bright for Gaia. They saturate the sensitive CCD detectors on-board of Gaia. The brightest object included in Gaia DR2 has a G magnitude (phot_g_mean_mag) of 1.71 but incompleteness is large until G ≈ 3. So practically, no star visible to the naked eye made it into Gaia DR1. More information on this topic can be found in the Gaia Data Release documentation or more specifically here.


Why do some sources have null 'parallax', 'pmra' and 'pmdec' values?

A five-parameter astrometric solution, including proper motions and parallaxes, has been computed for a sub-sample of about 1,3 billion sources of the total of 1,7 billion sources in Gaia Data Release 2; and has been computed for the TGAS sub-sample comprising circa 2 million sources in Gaia Data Release 1. More information can be found here.


Why are some parallaxes less than zero?

Negative parallaxes are caused by errors in the observations. Even if a negative distance has no physical meaning, there are a certain number of stars expected to have negative parallaxes just from an error propagation perspective. The negative parallax tail is a very useful diagnostic on the quality of the astrometric solution. Further details can be found here and here.


How can I extract Gaia data for my list of targets with RA and DEC?

There are two ways to do this:

  1. Basic: create a list of RA DEC coordinates (no header, one entry per line, separated by spaces) and upload in the GUI. See next FAQ for further details.
  2. Advanced (ADQL): follow the White Dwarfs Exploration tutorial here.


I have a problem uploading a list of targets in the simple form. Why?

The file to be uploaded to the "simple form" of the Gaia Archive online search tool requires a specific format. The description of the format can be found by hoovering the mouse over the "Select a file with Target Names" and clicking once the question mark appears.

The input file has no header and one entry per line. Three different name resolvers are tried: Simbad, NED and Vizier. An intermediate job with the retrieved source properties (name, coordinates, parallax and proper motion, if available) is created and cross-matched against the Gaia Archive. A mixture of object names and positions in the sky (see previous FAQ) is allowed.


The load_tables() method of astroquery.gaia.Gaia returns an SSL Error. What is going wrong?

This problem most likely originates from the way Python is installed. Native Python distributions (e.g. those downloaded from python.org) include an “Install Certificate” script that install a set of SSL certificates. However, this is not the case if the user is running the Python distribution that comes along with the Conda package. A workaround is to execute the following command in a terminal:

export SSL_CERT_FILE=/lib/pythonX.Y/site-packages/certifi/cacert.pem 

Where "pythonX.Y" should be changed by your running version of Python (e.g. "python3.8").


It is possible to delete all the jobs in my jobs list at once?

The simplest way to do this is to manually select all the jobs that are listed in the Archive and then press the "Delete selected jobs" button on the bottom right corner of the Archive webpage. An alternative to this potentially tedious task is to use the Astroquery.Gaia Python module (link).

The steps to do this are detailed below:

from astroquery.gaia import Gaia


jobs = [job for job in Gaia.list_async_jobs()]
# To print all the jobs owned by the user:
for job in jobs:

# To remove all the jobs at once:
job_ids = [job.jobid for job in jobs]


Is there a different behaviour when using integer or float division in the query?

Yes, when querying while using a division in the query, it makes a difference when one writes 1000/60 or 1000./60. The difference is that ADL does integer division in the first case and float division in the second case. Hence, results will be different!


More frequently asked questions on the Gaia Mission and Gaia Data Releases can be found here.



Obsolete FAQs


Why do I find non-compliant units in FITS headers?

The headers of FITS files of Gaia Data Release 1 as served from the Gaia Archive indeed contain constructs like TUNIT16 = 'Angular Velocity[mas/year]' which are not compliant with FITS standard 3.0. This was corrected for Data Release 2 (DR2). Gaia DR1 users can continue using the FITS files: their data contents are correct while the non-compliances in the headers may give warning messages (but not errors).


I use the SAMP at the job result but nothing happens. Why?

There may be several reasons for this. First of all the SAMP hub must be up and running. For instance, if TOPCAT is being used a SAMP hub is automatically started.

Then there is a known limitation of current SAMP standard related to the use of HTTPS connections. Current implementations of SAMP hubs does not support HTTPS connections. If you are using the Gaia Archive in HTTPS mode this can be the cause of the inability to connect to the SAMP hub. There are two possible workarounds for this.

First workaround: allow your browser to connect to insecure content. This will allow connections to the SAMP hub using HTTP from your secure Gaia Archive application.


Second workaround: use the Gaia Archive HTTP version. Be aware though that the latter implies a security risk as the communication between your browser and the Gaia Archive server is not being encrypted!



More frequently asked questions on the Gaia Mission and Gaia Data Releases can be found here.