Where can I find a catalog of all stars in the Milky Way?

  • Is there a catalog of all known stars or brightest stars in our galaxy? Preferably with some sort of galactic coordinates instead of just night sky coordinates (right ascension, etc.) I'm trying to make a model of the milky way, explorable with an oculus rift.

    @self He says, "known stars". It implies catalogued. Otherwise, you got a point

    What's the process like for cataloguing? Can we just go through celestial images taken from two locations, mark the stars and triangulate? Is this something that could potentially farmed out to Mechanial Turk?

    @Cheeku I missed that. :(

    @Axiverse Hard code the most known stars, and randomly/procedurally generate the rest. I hope to play you game someday.

  • Hipparcos, the predecessor to Gaia, has a dataset (http://cdsarc.u-strasbg.fr/viz-bin/Cat?I/239) with 3D positions for 100,000 stars. While we have much larger datasets of galactic stars, such as SDSS, finding the distances to stars is much harder. Parallax is the most precise technique for finding distances. Even for Gaia, we will only be able to measure distances to 10% precisions.

    There is a "galactic coordinate system", but that stil has the sun at the center and the line between the sun and the galactic center as its reference. There are calculators that can convert RA and Dec to galactic coordinates (e.g. http://python4astronomers.github.io/astropy/coordinates.html). You can use that and the distance to place them in space.

    It should be noted that any such catalog does not contain all stars in the galaxy. That would be impossible at the moment. However, existing catalogs represent best current efforts - so, in a way, that is the answer to the OP.

    The 10% number for Gaia is incorrect.

  • No, such a catalogues does not (yet) exist. There are two reasons.

    1 The Milky Way galaxy is about 20kpc (1pc ~= 3 lyr) across and only the very brightest stars are individually identifyable across such large a distance (such bright stars by their nature are very massive and hence young). Astronomers tend to cataloge stars by their apparent brightness, which for stars of identical luminosity declines as $1/d^2$ ($d$=distance). As a consequence, most catalogues contain only stars in the immediate galactic neighbourhood of the Sun. The Hipparcos catalogue (mentioned in another answer), for example, has most stars within a mere 100pc of the Sun.

    2 Obtaining distances for individual stars is inherently difficult, in particular the more distant the star in question is. Accurate distances for stare several kpc away can currently only be obtained by indirect methods applicable only to certain types of stars (such as RR Lyrae variables). The classical trigonometric parallax measurement for such distances, however, is subject of ESA's ongoing Gaia mission.

    ESA's Gaia satellite launched last year aims at cataloguing about $10^9$ stars across the Milky Way, including their velocity. The first preliminary versions of resulting catalogue, however, will still take some time to appear.

  • My understanding is that any star catalog today represents only such a very tiny and local part of the Milky Way, that you would have very little use of it for your purpose.

    In a year or two the Gaia space telescope will have mapped the one billion or 1% of the nearest and brightest stars in the Milky Way. Even then, to model a galaxy one needs other ideas than maps of individual stars.

    Is there a way to actively track Gaia's progress as they are mapping stars one by one?

    It seems that their first data release will be 22 months into the mission - so around end of 2015. Here's the link, still looking for a mailing list or intermediate data. Data Release Scenario

    Actually, more like beginning of 2017.

    @Axiverse that's not how Gaia works. It basically measures lots of angular distances between stars that are quite far apart in the sky. Over months it builds up so many of these that the data processing teams can use them to figure out where on the sky all the stars must be to make all the angles work out. Then when it looks at the same stars 6 months later, the changes tell them about the distances from Earth

  • Might not help the OP since the question is old, but I wanted to do something similar (use known star data for an n-body simulation). Basically, I realized there was no such catalog of stars for the Milky Way.

    The reason for this is also in other answers here, but simply put, there is a limit to how far our instruments (such as Gaia) can see due to 1) the technological and optical limits of our instruments 2) the galaxy itself blocking our view of the galaxy. I did a query for the maximum of "simple distance" (1/parallax) on the entire Gaia DR2 dataset and it seemed that the farthest stars are a little over 8k parsecs away, which is about the distance from Earth to Sag A* (the center of the galaxy).

    I downloaded a random sample of 3 million stars from Gaia DR2 and plotted them using OpenGL. Youtube reduced the video quality and my frame rate suffered while recording the video, but you can see it here. I used the galactic coordinate system here, where Earth is at the origin and the XY plane is the galactic plane. About halfway I move the camera from the origin to "outside" the galaxy.

    In the "outside" view, you'll see that there's no real "spiral arm pattern", just a blob. Apparently there is too much crap in space, and I suspect some filtering or other manipulation would be necessary to tease out the "spiral arm". However, the view from the origin is pretty good and what us Earthlings would "expect" to see. It's a rough (but 3d) version of a quality render done by the ESA.

    The spiral arm structures are only clearly seen if you correctly weight the stars by luminosity. But even then, it is probably smeared out by parallax uncertainties for individual stars. Also, did you correctly calculate the distances (i.e. not just inverting the parallax?) Have you checked out GaiaSky?

    No, I did it all "quick and dirty", especially the distance calculations (inverted parallax). I used solar luminosity L (which I think is already expressed in relation to the sun) to calculate bolometric magnitude Mb `-2.5*log10(L)`, and used Mb and parallax to calculate apparent magnitude. Since that is on a something like 1-16 scale, I clamped apparent magnitude to that range, then inverted to pseudo-normalize [16,1] to [0,1] "brightness". Then I multiplied brightness with a RGB representation of effective temp to get a final star color (as seen from Earth).

    Gaia Sky looks awesome. My little program is a crap version of that for sure, so I'm definitely doing a lot of calculations "wrong". It also just plots static position. I had to abandon the idea of using the gaia data directly for an n-body simulation since it's missing a lot of stars.

    Not quite right. The bolometric magnitude doesn't translate directly to a visual magnitude. You need to use bolometric corrections. You have worked out something that would be what you'd see if you could see at all wavelengths. I've yet to play much with GaiaSky, but I want to use it to produce some dynamic movies (fly-throughs) of some local clusters and also to show stuff moving around over thousands of years, all for projection in a small planetarium.

    You can see spiral structure if you do stuff like group the O and B stars into their associations and take their average parallax distance.

    Thanks for the tips Rob. I saw the bolometric correction stuff when I was reading about it, but at the time I just wanted to get something rendered and used the value as-is. I'll look into doing a better job on the magnitudes and the O & B stars and other things in a later phase. I'm not really that familiar with all the technical astronomy stuff, so it's a lot of reading and new concepts for me.

  • You might already have your answer at this point, but never mind, here is an up to date (2018) answer:

    The Gaia Data release is out (1 & 2).
    Here is a link to the download page of the Gaia archive data release 2: http://cdn.gea.esac.esa.int/Gaia/gdr2/gaia_source_with_rv/csv/

    If the link is dead or you want more, you can just type Gaia archive on a search engine and then go to downloads and you should find what you are searching for. You can also select only the ones that also have radial velocities if you want to include time evolution.

    For more precise requests, you can register on the Gaia archive (just need email address + name and takes 5 minutes) and you can then do queries to the database allowing you to filter the sources as you please.


    1. The distance measurements have about a 10% margin of error, but it should be the state of the art for the time being.

    2. There are plenty of parameters more than just the astronomic parameters (position and velocity), that you can use for rendering your application as accurately and interestingly as you like.

    Gaia does not measure all the stars in the galaxy, as has been pointed out in other answers. The distances aren't just accurate to 10%, it is much more complicated than that.

    @RobJeffries you are indeed right, it is not all the stars in the galaxy but a good number considered the desired application. To my knowledge it is one of the best catalogue for doing what **Axyverse** wanted to achieve. You are right also for the error, it is much more complicated, depending on the characteristics of the star (e.g. how much it moves, or its change in colour that Gaia doesn't take into account to my knowledge), but the 10% is to give a approximative idea, since I guess for an VR application I am not sure how much exactitude is needed and a global idea might be sufficient.

  • Galaxy Map has detailed information on 5000+ stars.

    You can download the exel spreadsheet here

    Don't know if it helps. Good luck ;)

License under CC-BY-SA with attribution

Content dated before 7/24/2021 11:53 AM