I'm looking for some sort of boundary data to be able to render the milky way on a star map, as visible from Earth. Something that looks like this:

star map

For that, I need something like a collection of RA hours and DEC degrees of the "boundary points" of what's visible from Earth (technically the Galactic Center), possibly with proper motion too. I'm not looking for precise luminosity data or anything like that, just the points of the blob on the sky that most resembles the Milky Way's shape and position from the Earth. It's important that I want to render the sky map for any given surface point on Earth, for any given time (within the last 100 years at least).

Do you know of a database like that? I've been looking on VizieR but I couldn't find what I was looking for.

  • 2
    $\begingroup$ There are a lot of panoramic photographs of the milky way view out there - I wonder if it's worth hand-tracing a few outlines from these and storing RA-Dec for the line segments... $\endgroup$
    – Andy
    Commented Sep 6, 2016 at 11:15
  • $\begingroup$ I guess as long as I know the exact time and place the photograph was taken it's not a bad idea. I'd need to cover the whole surface the the Earth though (at least so that I can calculate the look of the Milky way from anywhere) $\endgroup$
    – gphilip
    Commented Sep 6, 2016 at 11:27
  • $\begingroup$ I was thinking these (probably public domain) photos show plenty of stars too, so once you identify the stars it might be possible to use their positions on the photo to identify RA/Dec for the milky way clouds, etc too. (One problem is I don't know if such pics have any distortions caused by the projection method chosen. That would have to be accounted for.) $\endgroup$
    – Andy
    Commented Sep 6, 2016 at 11:35
  • $\begingroup$ By the way here is an interesting method (collecting lots of stars from an Infra Red survey) though it doesn't present the usual cloudy appearance. $\endgroup$
    – Andy
    Commented Sep 6, 2016 at 11:42
  • $\begingroup$ Have you already tried Sky within Google Earth or some Sky map app on the smartphone or tablet to locate the current position of the Milky Way? $\endgroup$
    – ott--
    Commented Sep 6, 2016 at 20:39

5 Answers 5


There is a very nice project called d3-celestial by Olaf Frohn on github. In contains a data file describing the Milky Way as polygons, see here. A demo showing this Milky Way can be found here. And even better, the source for this data is cited, pointing to the Milky Way Outline Catalog by Jose R. Vieira.

Depending on your project, the json format from d3-celestial might be easier to read than the one from Jose R. Vieira. Note that you don't have to worry about these "contours" moving on a time scale of hundred years, but this is another question.

  • $\begingroup$ Cartes du Ciel also appears to use the Vieira outline data. $\endgroup$
    – Mike G
    Commented Sep 9, 2016 at 17:45

Following @Petoetje59's idea of using the data in the "mw.json" file I was able to add a fine Milky Way contour rendition to my proper motions planisphere project.

I just took all the coordinate pairs en masse, eliminated all the brackets and code, added line feeds at the end of each pair and pasted them on a spreadsheet, where I could make the stereographic projection of all the points.

Finally I plotted the projected points in Inkscape, separated each individual contour polygon and stacked all of them; by coloring them in transparent blue I got that nice layered look.

Thanks very much!

Milky Way Contour


From memory (I don't have a copy here right now) the HNSky package used to have a hand-drawn "supplement" file for the milky way. It contained a lot of data points arranged as RA & Dec points and was extracted from photographs, with a bit of correction and scaling.

I suggest you load and install one of the older versions still on HN Sky's page, such as 3.0.0, and look for the supplement file. (The newer version 3.2.3 does not seem to include this file.)

This is only position data - no proper motion, etc. as the question mentioned. So only a partial answer.

(Note: You might need to ask the authors of that file for permission if you're going to release software based on this. No harm in looking at it for purely personal interest of course.)


The planetarium software, Guide, to be found at ProjectPluto.com, has a map of the Milky Way. This is how the dataset is described

"The Nebula Databank was compiled by Eric-Sven Vesting to evade the problems that came with earlier bright nebula databases. For example, previous versions of Guide gathered nebula data from five separate catalogs. There were few cross-indexes from one catalog to another, and no way to indicate that one designation applied to a part of larger area with a different designation. Also, brightness levels were applied in an inconsistent manner at best. The Nebula Databank contains explicit links between the various nebula catalogs, enabling Guide to show all designations for a given object and to avoid drawing some objects twice (if they appeared in separate catalogs under different names). Also, Eric-Sven Vesting created the nebula isophotes used by Guide to indicate the shapes of most prominent nebulae. Better positional data was generated, usually by comparing catalog positions to actual RealSky images."

enter image description here (I hope I have not violated any copyright on this).

It's not clear where that database is, but perhaps you could contact either ProjectPluto or Eric-Sven Vesting directly.

I have no affiliation or connection, financial or otherwise, with this software.


I figured out the meaning of the GeoJSON data in "mw.json" (D3-Celestial, Olaf Frohn) describing the Milky Way boundary.

There are 5 groups (brightness levels, 1 darkest to 5 brightest) of coordinates describing polygons that delineate the Milky Way on the sky dome : 10 polygons for brightness level 1, 113 at level 2, 46 when level 3, 27 with level 4, and 6 polygons for level 5.

The coordinates [longitude,latitude] are in degrees with up to 3 decimals.

Right ascension ($\alpha$, [0°,360°[) equals positive 'longitude', otherwise add 360° to the 'longitude' to obtain $\alpha$. Divide $\alpha$ (in degrees) by 15 to obtain it in hours.

Declination ($\delta$, [+90°,-90°]) always equals the 'latitude'.

Equatorial coordinates $\alpha$ and $\delta$ are given for epoch J2000.

The plotted Milky Way should look like this.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .