Data releases

This catalogue is corrected continuously, which is the opposite of what a citation needs. A release is one state of it, frozen: a date, a count of every row, the licence and retrieval date of every catalogue behind it, and a download that will not change again. Cite a release rather than the site — the API names one in every response, so a number quoted today can still be found in five years.

[

DR1

] 2026-08-21 · NOT YET DEPOSITED

First public data release

Every distance carries the method that produced it, every modelled stellar parameter carries the quality cut that admitted it, and every galaxy distance is computed under one stated cosmology.

OBJECTS
223 448
DESIGNATIONS
520 680
RELATIONS
3396
UPSTREAMS
16

STAR 124 271 · GALAXY 64 158 · ASTEROID 22 343 · EXOPLANET 6336 · COMET 4072 · CLUSTER 935 · MOON 459 · NEBULA 433 · OTHER 419 · BLACK_HOLE 14 · PLANET 8

DOWNLOAD THE FROZEN BUNDLE →WHAT EVERY COLUMN MEANS →OBJECTS · DESIGNATIONS · RELATIONS · SOURCES · SHA256SUMS
WHAT IT SHOULD HASH TO
  • spacecatalog-dr1-objects.csv.gz27 573 650 bytes
    630ba7ae703659f0c4fe67480d11dd1f2c7eafd89bb2443361d68d9e91321398
  • spacecatalog-dr1-designations.csv.gz3 531 501 bytes
    1ea90d8d329d8b5b03f84fa15846a9ca2ecada5825dfd984fd0bc156b17114e6
  • spacecatalog-dr1-relations.csv.gz37 530 bytes
    b2e41ea6a22427a53abb7e7b09cb4bd875892ac5302e2e233e4dfd2a0e6382ff
  • spacecatalog-dr1-sources.csv2496 bytes
    8f9c4d6441d1ed5c3ba38c35164ae275726167b3609df8e7eb2a1732dcb57ea4
  • spacecatalog-dr1-source_disagreements.csv8628 bytes
    af601cb342ef877bde356bcc2ffff35dda5d0197f8167188482b3f6fd9e4b71b
  • LICENSE11 357 bytes
    c71d239df91726fc519c6eb72d318ec65820627232b2f796219e87dcf35d0ab4
  • RELEASE.md2993 bytes
    44e3ae9ce537c5dbc7df1859b20f27ab6bb63b61ca561e6c05bdb100f8a04859
  • spacecatalog-dr1.tar.gz31 091 613 bytes
    0b231d69a6d8e60ebaebfb7a88238905a94cf82cf8775e83f375a554c5abb4f0

DR1 is the first state of the catalogue meant to be cited. There is no release before it, so what follows describes what a row is rather than what changed.

Every object names the catalogue it came from and the licence that catalogue publishes under, and every distance names the method behind it. For an exoplanet system that method is the archive column and the paper the archive credits for it — which for 280 microlensing systems is a Bayesian estimate from a galactic model with no parallax measured anywhere in it, and not the parallax the label used to claim.

Modelled stellar parameters are gated rather than republished. Gaia's photometric temperatures, radii, masses and ages are withheld from 4,592 stars whose astrometric solution failed the standard binarity test (RUWE > 1.4, Lindegren et al. 2021), and the 17,874 kept values that fall in a regime the model is known to handle poorly — too hot, or classified as a pair — say so on the row.

Doubt is carried in the numbers rather than in prose. Published uncertainties are stored as [minus, plus] pairs; a distance derived from a redshift carries the redshift's own measurement error together with the ±4% the expansion rate is uncertain by, added in quadrature. For the 31,510 galaxies whose redshift is photometric rather than spectroscopic, that bar is twelve times wider than what earlier versions of these pages showed.

Where a redshift is small enough that the choice of reference frame decides whether there is a Hubble flow to measure at all, no distance is given in either frame: 93 galaxies, several of them well known, carry no distance here rather than one that was right by accident.

What it was built from

Each date is when that catalogue was last downloaded for this release, not when the release was assembled.

What the checks found

95 of 223 448 rows carry a data-quality note — impossible density (95) — and nothing else fires. A release where one of them matches more rows than the last accepted count does not get built.

Two catalogues describe one object differently beyond the tolerance for that quantity in 82 places. Those are listed in the bundle rather than resolved in silence: catalogues disagreeing is the normal condition of astronomy, and which one is right is a question this release records rather than answers.

Conventions this release is fixed to

Both in one line each; the arithmetic behind them, and what was done to every upstream’s numbers, is on the methods page.

  • Cosmology

    Redshift distances are computed at H₀ = 70 km/s/Mpc, flat ΛCDM with Ωₘ = 0.3153, from a redshift in the CMB rest frame (Planck 2018 dipole). The ±4% the expansion rate is uncertain by is common to all of them, so it does not cancel against a distance measured any other way.

  • Positions

    ICRS, epoch J2000.0 in every row.

[

LICENCE

] 11 DISTINCT TERMS

The data is not under one licence. Every row names its catalogue in the source column and that catalogue’s terms govern it — the 11 sets of terms in this release are listed against the catalogues they cover above, so one summary line here would be wrong about most of it.

3 of them are share-alike, reaching anything derived from the rows they cover. If you redistribute a subset, check which catalogues it drew from.

[

CITING A RELEASE

] DR1

No DOI has been minted for this release yet, so the citation points at this page. It will gain a DOI without the release changing: a release is its rows, and depositing the bundle does not move one of them.

SpaceCatalog DR1: First public data release. 2026. https://spacecatalog.org/releases
@misc{spacecatalog_dr1,
  title  = {SpaceCatalog DR1: First public data release},
  year   = {2026},
  url    = {https://spacecatalog.org/releases},
  note   = {Data release, retrieved from https://spacecatalog.org}
}

This does not discharge the upstream attribution: any row you quote names the catalogue it came from, and the share-alike ones ask to be named where their rows are used.