You are here

Kevin Jardine's blog

Making a galaxy map, Part 2: Slicing

Submitted by Kevin Jardine on 19 June, 2018 - 07:57
Hot star density slice near the galactic plane
An example hot star density slice (in this case near the galactic plane).

In order to apply the isosurface mapping technique, we must convert a point cloud, such as the set of about 340 thousand hot O, B and early A class stars I described in my last blog post, into a scalar field, a real valued function defined in principle for every point in euclidean space (or at least in a defined grid). This scalar field is in some way a measure of star density. We can then extract isosurfaces of constant density from this scalar field, much as contour lines are lines of constant elevation on a topographic map of mountainous terrain on Earth.

In order to make the necessary calculations manageable on my 16 Gb workstation, I placed the density values into 3x3x3 parsec bins.

I represented the cylinder mentioned in my last blog post (with a radius of 3000 parsecs and height 600 parsecs above and below the galactic plane) as a subset of a 6000 x 6000 x 1200 pc box and reduced this into 2000 x 2000 x 400 bins (with dimensions 3x3x3 pc) in a numpy array.

I then counted the number of stars in each bin. This created a scalar field of sorts, but a very discontinuous one.

I borrowed the gaussian smoothing technique mentioned in H. Bouy and J. Alves 2015, "Cosmography of OB stars in the Solar neighbourhood". For the map I chose a a gaussian sigma of 15 parsecs. The choice of a gaussian sigma is as much an aesthetic choice as it is a scientific one. The higher the sigma, the less detailed the map. A low sigma results in a chaotic map. A high sigma results in a smooth map with limited detail. I experimented with several values before settling on 15 parsecs.

Fortunately, the Python library scipy.ndimage.filters has an appropriate gaussian_filter function that could be applied directly to the numpy array.

In addition, for ease in future processing, I wanted to ensure that the values of the density function ran between 0 and 1.

There are numerous ways to normalize the density values in this way (for example dividing by the maximum value). The option I chose is to use this variation of the sigmoidal logistic function:

f(x) = 2/(1+exp(-kx)) - 1

which ranges between 0 and 1 in an S-shaped curve as x ranges between 0 and infinity.

This sigmoidal value stretching / normalization is often used in image processing to continuously compress data with very high and low values into a smoother middle range. As with the gaussian sigma value, the logistic k value is partly an aesthetic choice. A low k value pushes the function towards 0. A high value pushes the function towards 1. After experimenting with various values for k, I selected 300. This seemed to more evenly spread the possible density values between 0 and 1.

The scipy.special library has an expit function that can be applied to numpy arrays.

I then sliced the cylinder values in the numpy box from top to bottom in 3 pc slices and output the results as coloured 8 bit images (for debugging the sigma and k values) and as 16 bit pgm values (needed for isosurface generation). One of the coloured slice images is at the top of this blog post.

Once the slices were done, I could move on to the next stage: meshing.

Making a galaxy map, Part 1: Data

Submitted by Kevin Jardine on 18 June, 2018 - 08:15
ESA Gaia ADQL advanced search
The public archive site can be used to download Gaia DR2 data.

Last week I released a map of the Milky Way within 3000 parsecs (10 thousand light years) of the Sun made using data from Gaia DR2 and several other sources as described on my previous blog post.

You can see a face-on web version here or view the full 3D meshes in the latest version of Gaia Sky.

This week I'm starting a series of blog posts to explain in detail how I created the map. I hope that these posts will be useful for others who might want to try the density isosurface technique to create their own maps. At the end of this series I'll link to relevant resources, including the map meshes, SVG and Blender files, and Python source code.

The Gaia DR2 archive provides parallaxes for more than a billion stars, but I used only a tiny fraction of these to create the actual map. First, I selected only the stars with the most accurate parallaxes in the Gaia DR2 data set, the 72 million stars with error/parallax < 0.1.

There is a field in the Gaia DR2 database that is the reciprocal of this value, parallax_over_error. So in practice I was selecting the stars with parallax_over_error > 10.

The main difficulty with extracting this data set from the Gaia archive was the time. For the main archive, there is a maximum query limit of 30 minutes and at the time I was using the archive, selecting the data was slow. I ended up splitting the data selection into 28 queries based on the parallax value. A sample query is shown in the image at the top of this blog post. It took most of a day to get all the data. I do wonder if the Gaia database administrators might consider allowing faster downloads for commonly used data subsets.

After acquiring the data, I applied a few quality cuts and selected only about 400 thousand very hot O, B and early A class stars using a colour index filter with colour index < 0 and added an absolute magnitude cut (absolute magnitude < 7) to filter out dim but hot non-main sequence stars like white dwarfs. (Actually the hot stars I was mapping all have an absolute magnitude much brighter than 7. I chose that number to compensate for some dust extinction as well.)

The Gaia DR2 database includes a raw colour index value bp_rp, and a second colour excess value, e_bp_min_rp_val, that attempts to correct for dust reddening.

So we can define colour index = bp_rp - e_bp_min_rp_val.

These values are a relatively simple estimate of what Gaia will ultimately be capable of - more accurate colour values will be provided in Gaia DR3.

Filtering to these hot, young stars was an essential part of making the map possible. Cooler, older stars tend to drift randomly from their sites of formation. I tried a number of experiments to map cooler stars without a great deal of success - they were either present in a large number of random clumps or, at lower density levels, just a few large areas.

In contrast, hot young stars are largely gathered in dense associations, mostly near the galactic plane. Their distribution, especially those regions containing the ultra hot O-B3 class ionizing stars, form a distinct (and hence mappable) pattern.

After the colour selection, I further filtered to all stars within a cylinder with a radius of 3000 parsecs and a height of 600 parsecs above and below the galactic plane.

Why the 3 kpc limit? The main reason is that stars with very accurate parallaxes (err/plx < 0.1) start to fade out beyond 3.6 kpc. I definitely wanted to include the Carina nebula, which is more than 2 kpc away. The further I extended the map, the less accurate the star positions became, and also, more pragmatically: most of the structures (eg. OB associations or HII regions) that astronomers have named turn out to lie within 3 kpc. So a 3 kpc limit allowed the most accurate map with names for most of the major regions.

The 600 parsec height limit was a bit arbitrary and was mostly to reduce memory consumption. This did result in some truncation for a few very low density regions. An interesting alternative map might ignore the thin disk and look for density structures in the thick disk and halo.

This distance cut gave me a final data set of about 340 thousand stars, ready for the next stage of density isosurface generation - slicing.

A map of the Milky Way out to 3000 parsecs (10 thousand light years)

Submitted by Kevin Jardine on 14 June, 2018 - 08:51
Face-on map of Milky Way
A detail from the face-on map of the Milky Way. The full map poster is is available on the web here.

Sometimes dreams do come true. Today I can announce a detailed map of the Milky Way out to 3000 parsecs or about 10 thousand light years from the Sun.

I developed this map with help from scientists with the European Space Agency's Gaia mission and researchers at Leiden and Heidelberg universities. It includes star density isosurfaces mapping the major concentrations of the hotter O, B and A class stars in the Gaia DR2 release, about 5000 extremely hot ionizing stars, dust clouds and HII regions. Even better, it is available both in a face-on form viewed from above the Milky Way and in a true 3D version in the latest version of Gaia Sky.

A more detailed description of the map is up at the ESA Gaia mission website here.

A printable pdf is available here. (Warning, even at A0 size, the smallest labels are tiny. Needs sharp eyes, a magnifying glass or printing at a custom size with a width greater than one metre.)

A zoomable pannable web version is available here.

There is a gray background in the face-on versions of the map that shows the approximate locations of the inner and outer galaxy.

You can download the latest version of Gaia Sky with 3D mesh support here.

A Youtube trailer showing a flight through the map, animated in Gaia Sky, is available here:

Many people have helped make this map possible. I'd like to mention a few now.

Anthony Brown, a professor at Leiden University and the chair of Gaia's Data Processing and Analysis Consortium (DPAC), has been an enthusiastic supporter of this project and was always available with suggestions and comments.

Toni Sagristà Sellés, the principal developer of Gaia Sky, added numerous features to Gaia Sky to support visualizing and animating the 3D meshes.

Stefan Jordan, a professor at Heidelberg University, the sponsor of Gaia Sky, and Manager for Outreach and Education for DPAC, has also provided support and encouragement.

And at the European Space Agency itself, Jos de Bruijne and Tineke Roegiers, two Gaia Mission scientists, provided advice and support at key points in this project.

Finally, I'd like to thank Bob Benjamin, professor at the University of Wisconsin - Whitewater, who has long encouraged my interest in galactic cartography and made useful suggestions for positioning the HII regions on the current map.

The map depends upon data from many researchers. Here are some key credits:

Gaia Data Release 2 data as available from the Gaia Archive.

The isosurface technique was first pioneered for Hipparcos by H. Bouy and J. Alves, as explained in "Cosmography of OB stars in the Solar neighbourhood".

The map also includes dust density isosurfaces computed by dust extinction values provided in a preprint article "3D maps of interstellar dust in the Local Arm: using Gaia, 2MASS and APOGEE-DR14" by Lallement et al.

Positions for OB associations are based on the computed median distances to their members in the data set as described by R.M. Humphreys and D.B. McElroy in "The initial mass function for massive stars in the Galaxy and the Magellanic Clouds".

Positions for HII regions are computed using the median distances for known ionizing stars as described here.

The map also includes an overlay of about 5000 known ionizing stars. Distances for these stars are derived from a catalogue associated with a preprint article by C.A.L. Bailer-Jones et al in "Estimating distances from parallaxes IV: Distances to 1.33 billion stars in Gaia Data Release 2".

Of course all errors are mine. As with any map in the early days of exploration, it is bound to have missing sections and regions with incorrect data. But I think that it is a good start. I'm excited about the current map and even more looking forward to the many new features and improvements that we'll be adding in the future.

I've been running this site for almost 14 years now but today feels like a new beginning.

Some Like It Hot, Part 5, A revised face-on map within 3kpc using Gaia DR2

Submitted by Kevin Jardine on 12 May, 2018 - 16:38
Face-on map of Milky Way
A face-on map of the Milky Way within 3kpc (about 10 thousand light years). The full revised map is is available as a PDF here.

I have revised the face-on map I published in my last blog post. As before, this is based on the Humphreys/Blaha data set of ionizing and highly luminous stars and a (now expanded) list of ionizing stars for HII regions. For this revision, I used the distance estimates for Gaia DR2 stars provided by Bailer-Jones and his colleagues at Germany's Max Planck Institute for Astronomy. These distances provide more accurate distance estimates for many stars than a simple reciprocal calculation, and allow distances to be estimated for a much larger set of stars.

In order to provide an additional check on the HII region distances, I have compared the Bailer-Jones distances for the ionizing stars with existing photometric and kinematic distance estimates for the associated HII regions. If a photometric or kinematic distance estimate provided an error range, I used that, or otherwise used a range of +/- 25% of the published value.

If an ionizing star had a Bailer-Jones distance outside the error range for any published distance estimate for the associated HII region, I considered it an outlier and filtered it out. I then computed a filtered distance for the HII region as the median distance for the filtered ionizing stars. If 50% or more of the ionizing stars for an HII region made it through the filtering process, I used the filtered distance value for the revised map. If less than 50% of the ionizing stars for an HII region made it through the filtering process, I excluded the HII region from the map.

For a small number of HII regions with very bright ionizing stars I used Hipparcos stars with err/plx < 0.2 when Gaia DR data was not available. I used the reciprocal method to compute the distance and then applied the same filtering process described above.

Fortunately, almost all the major HII regions within 3kpc made it through the filtering process. The ones that did not were mostly small regions with few known ionizing stars. I have added a table below listing the HII regions that had at least one ionizing star with a Bailer-Jones distance compatible with a previously published HII region distance. I provided links to the relevant distance papers in the last column of the table.

The full revised map is is available as a PDF here. You will need to zoom into about 200% in the PDF to read the labels for each HII region.

You can also see the list of ionizing stars I used for the HII regions.

Region All distance All star count Filtered distance Filtered star count Percentage Sources
Sh 2-275 1508 6 1508 6 100% [1], [2]
Gum 38a 2574 4 2574 4 100% [1], [2], [3]
RCW 14 2778 4 2778 4 100% [1]
RCW 32 962 4 962 4 100% [1], [2]
Sh 2-157 3209 4 3209 4 100% [1], [2], [3]
Sh 2-299 4182 4 4182 4 100% [1], [2]
Sh 2-300 4182 4 4182 4 100% [1], [2]
Sh 2-306 4959 4 4959 4 100% [1], [2], [3]
DWB 78 1652 3 1652 3 100% [1]
Gum 21 2228 3 2228 3 100% [1]
RCW 104 3505 3 3505 3 100% [1]
Sh 2-150 982 3 982 3 100% [1], [2]
Sh 2-205 1141 3 1141 3 100% [1], [2], [3], [4]
Sh 2-219 4025 3 4025 3 100% [1], [2], [3], [4]
Sh 2-252 2174 3 2174 3 100% [1], [2], [3], [4]
Sh 2-254 2376 3 2376 3 100% [1], [2], [3], [4]
Sh 2-255 2376 3 2376 3 100% [1], [2], [3], [4]
Sh 2-256 2376 3 2376 3 100% [1], [2], [3], [4]
Sh 2-257 2376 3 2376 3 100% [1], [2], [3], [4]
Sh 2-258 2376 3 2376 3 100% [1], [2], [3], [4]
Ge 129 1999 2 1999 2 100% [1]
RCW 113 1545 2 1545 2 100% [1], [2], [3]
RCW 117 1470 2 1470 2 100% [1]
RCW 19 4091 2 4091 2 100% [1], [2]
RCW 41 2027 2 2027 2 100% [1]
RCW 47 2637 2 2637 2 100% [1], [2], [3], [4]
RCW 60a 2674 2 2674 2 100% [1]
RCW 75 2404 2 2404 2 100% [1], [2]
RCW 82 3280 2 3280 2 100% [1], [2]
RCW 88 2348 2 2348 2 100% [1]
Sh 2-139 3453 2 3453 2 100% [1], [2], [3], [4]
Sh 2-142 2964 2 2964 2 100% [1]
Sh 2-184 2306 2 2306 2 100% [1], [2]
Sh 2-217 4025 2 4025 2 100% [1], [2], [3], [4]
Sh 2-25 1195 2 1195 2 100% [1]
Sh 2-29 1072 2 1072 2 100% [1]
Sh 2-307 4054 2 4054 2 100% [1]
Sh 2-309 4959 2 4959 2 100% [1], [2], [3]
Sh 2-311 6450 2 6450 2 100% [1], [2]
Sh 2-34 1602 2 1602 2 100% [1]
Sh 2-4 3295 2 3295 2 100% [1], [2], [3]
Sh 2-92 3795 2 3795 2 100% [1], [2]
BFS 28 915 1 915 1 100% [1]
BFS 54 1514 1 1514 1 100% [1]
BFS 63 1090 1 1090 1 100% [1]
DWB 155 1659 1 1659 1 100% [1]
Ge 130 2910 1 2910 1 100% [1]
Gum 31 3289 1 3289 1 100% [1], [2], [3]
H 36 2773 1 2773 1 100% [1]
L 4 4455 1 4455 1 100% [1]
RCW 107 1101 1 1101 1 100% [1], [2]
RCW 108 1049 1 1049 1 100% [1], [2]
RCW 110 1900 1 1900 1 100% [1]
RCW 114 1154 1 1154 1 100% [1]
RCW 119 1749 1 1749 1 100% [1], [2]
RCW 125 1433 1 1433 1 100% [1], [2]
RCW 51 2566 1 2566 1 100% [1], [2], [3]
RCW 71 1608 1 1608 1 100% [1]
RCW 94 1779 1 1779 1 100% [1]
Sh 2-117 668 1 668 1 100% [1]
Sh 2-13 1039 1 1039 1 100% [1]
Sh 2-134 612 1 612 1 100% [1]
Sh 2-135 2252 1 2252 1 100% [1]
Sh 2-15 1249 1 1249 1 100% [1], [2]
Sh 2-155 831 1 831 1 100% [1], [2], [3]
Sh 2-158 2671 1 2671 1 100% [1], [2], [3], [4]
Sh 2-160 912 1 912 1 100% [1], [2]
Sh 2-162 2495 1 2495 1 100% [1], [2], [3]
Sh 2-163 3473 1 3473 1 100% [1], [2]
Sh 2-164 2840 1 2840 1 100% [1], [2], [3], [4]
Sh 2-165 1715 1 1715 1 100% [1], [2], [3]
Sh 2-168 2705 1 2705 1 100% [1], [2]
Sh 2-170 1887 1 1887 1 100% [1], [2]
Sh 2-173 3021 1 3021 1 100% [1], [2], [3], [4], [5]
Sh 2-175 2121 1 2121 1 100% [1]
Sh 2-180 4732 1 4732 1 100% [1], [2], [3]
Sh 2-185 168 1 168 1 100% [1], [2]
Sh 2-198 2176 1 2176 1 100% [1], [2], [3]
Sh 2-206 3135 1 3135 1 100% [1], [2], [3], [4]
Sh 2-207 3532 1 3532 1 100% [1], [2]
Sh 2-22 1082 1 1082 1 100% [1]
Sh 2-223 6602 1 6602 1 100% [1]
Sh 2-227 4259 1 4259 1 100% [1], [2], [3]
Sh 2-229 401 1 401 1 100% [1], [2]
Sh 2-232 2068 1 2068 1 100% [1], [2]
Sh 2-235 1651 1 1651 1 100% [1]
Sh 2-237 4139 1 4139 1 100% [1]
Sh 2-241 5258 1 5258 1 100% [1], [2], [3], [4]
Sh 2-242 1967 1 1967 1 100% [1], [2]
Sh 2-247 2027 1 2027 1 100% [1], [2], [3]
Sh 2-253 4540 1 4540 1 100% [1], [2]
Sh 2-261 2041 1 2041 1 100% [1], [2], [3]
Sh 2-263 529 1 529 1 100% [1]
Sh 2-264 279 1 279 1 100% [1], [2]
Sh 2-266 5523 1 5523 1 100% [1]
Sh 2-268 1141 1 1141 1 100% [1], [2]
Sh 2-27 181 1 181 1 100% [1]
Sh 2-28 2099 1 2099 1 100% [1], [2]
Sh 2-280 1488 1 1488 1 100% [1], [2]
Sh 2-283 6794 1 6794 1 100% [1], [2], [3]
Sh 2-298 4319 1 4319 1 100% [1]
Sh 2-302 2135 1 2135 1 100% [1], [2], [3]
Sh 2-32 1590 1 1590 1 100% [1]
Sh 2-44 1823 1 1823 1 100% [1]
Sh 2-46 1421 1 1421 1 100% [1]
Sh 2-54 1976 1 1976 1 100% [1], [2]
Sh 2-61 2390 1 2390 1 100% [1], [2]
Sh 2-7 150 1 150 1 100% [1]
Sh 2-72 1693 1 1693 1 100% [1]
Sh 2-80 6203 1 6203 1 100% [1]
Sh 2-82 846 1 846 1 100% [1]
Sh 2-85 484 1 484 1 100% [1]
Sh 2-90 2266 1 2266 1 100% [1]
Ori-Eri Bubble 396 15 396 15 100% [1]
Sh 2-119 787 6 877 5 83% [1], [2], [3]
Sh 2-131 956 17 953 14 82% [1], [2]
RCW 53 2503 51 2405 41 80% [1]
Sh 2-8 1754 20 1734 16 80% [1], [2], [3]
RCW 99 2396 5 2428 4 80% [1], [2], [3]
Sh 2-137 470 5 705 4 80% [1], [2], [3]
Sh 2-199 2039 5 2028 4 80% [1], [2], [3]
Sh 2-236 3785 5 3798 4 80% [1], [2], [3], [4], [5], [6]
Sh 2-297 1125 5 1103 4 80% [1], [2]
Sh 2-11 1617 28 1577 21 75% [1], [2], [3], [4]
Sh 2-273 772 12 764 9 75% [1]
Sh 2-230 2811 4 2820 3 75% [1]
Sh 2-296 1138 4 1133 3 75% [1], [2]
RCW 62 2527 7 2519 5 71% [1]
Sh 2-190 2171 9 2072 6 66% [1], [2], [3]
Sh 2-49 1961 9 1778 6 66% [1], [2]
DWB 68 1023 3 1026 2 66% [1]
Ge 68 2820 3 2820 2 66% [1]
Gum 37 3254 3 2953 2 66% [1], [2], [3]
m Cen nebula 2314 3 2183 2 66% [1]
RCW 102 3723 3 4055 2 66% [1], [2]
RCW 130 1227 3 1108 2 66% [1], [2]
RCW 33 1011 3 1011 2 66% [1], [2]
RCW 35 1680 3 1857 2 66% [1], [2]
Sh 2-295 1283 3 1269 2 66% [1]
Sh 2-31 1072 3 1072 2 66% [1]
Sh 2-88 2084 3 2005 2 66% [1], [2], [3]
Sh 2-45 1664 11 1678 7 63% [1], [2]
Sh 2-101 1936 5 1993 3 60% [1], [2], [3]
Sh 2-140 1069 5 937 3 60% [1], [2], [3]
Sh 2-30 1140 5 1529 3 60% [1], [2]
Sh 2-281 404 5 417 3 60% [1], [2], [3]
RCW 55 4606 4 4606 2 50% [1], [2], [3]
RCW 61 4097 4 2743 2 50% [1], [2]
Sh 2-108 1695 4 1478 2 50% [1], [2]
Sh 2-145 1089 4 982 2 50% [1], [2]
Sh 2-35 1668 4 1668 2 50% [1], [2]
Gum 12 337 2 342 1 50% [1]
RCW 129 180 2 145 1 50% [1]
RCW 146b 2770 2 1558 1 50% [1], [2]
RCW 38 2125 2 1740 1 50% [1]
RCW 48 2989 2 1960 1 50% [1]
RCW 68 2259 2 1924 1 50% [1]
RCW 83 2069 2 1881 1 50% [1]
RCW 85 1916 2 1605 1 50% [1], [2]
Sh 2-1 1445 2 179 1 50% [1]
Sh 2-220 758 2 381 1 50% [1]
Sh 2-289 2814 2 5534 1 50% [1]
Sh 2-292 731 2 1333 1 50% [1]
Sh 2-301 3790 2 4451 1 50% [1]
Sh 2-310 4000 2 1508 1 50% [1], [2]
Sh 2-37 1386 2 1536 1 50% [1], [2], [3]
Sh 2-64 681 2 566 1 50% [1]
Sh 2-87 1555 2 2310 1 50% [1], [2], [3]
Sh 2-171 1107 4 1039 2 50% [1]
DWB 34 1693 7 1693 3 42% [1]
Sh 2-41 1875 7 1875 3 42% [1]
Sh 2-129 1256 5 834 2 40% [1]
Sh 2-202 1029 5 819 2 40% [1], [2]
Sh 2-284 3869 13 4663 5 38% [1], [2], [3], [4], [5]
Gum 34b 6749 3 3450 1 33% [1]
RCW 106 2088 3 2088 1 33% [1]
RCW 20 4448 3 3607 1 33% [1]
RCW 49 3208 3 5805 1 33% [1]
RCW 50 2684 3 4052 1 33% [1]
RCW 74 2252 3 3707 1 33% [1], [2]
Sh 2-115 1704 3 2701 1 33% [1], [2], [3]
Sh 2-132 4542 3 4146 1 33% [1]
Sh 2-249 2384 3 2012 1 33% [1], [2]
RCW 105 4094 5 2115 1 20% [1], [2]

Some Like It Hot, Part 4: A face-on map within 3kpc using Gaia DR2

Submitted by Kevin Jardine on 26 April, 2018 - 17:39
Face-on map of Milky Way detail
A detail from a face-on map of the Milky Way within 3kpc (about 10 thousand light years). The full map is is available available as a PDF here.

Now that Gaia DR2 is available, it is possible to do an enormous amount of mapping and stellar analysis, much of which is announced on Twitter via the #GaiaDR2 hash tag.

I have used the two data sets I mentioned in my last two blog posts (The Humphreys/Blaha catalog of luminous stars and the expanded Avedisova catalog of ionizing stars for HII regions) to create a simple map of the Milky Way out to 3kpc (about ten thousand light years). It is available as a PDF here.

There is a lot of detail on the map, so you will need to zoom into 100% to see it. In general I used the Sharpless, Gum, Georgelin and RCW names for the HII regions, following Avedisova, which are not necessarily the most commonly used names today. At some point I will do a version with expanded labels.

Blue dots are OB and Wolf Rayet stars. Magenta dots are cooler stars. The yellow dot at the centre is the Sun. Red circles are HII regions.

I filtered the GaiaDR2 cross matches to only select stars with err/parallax < 0.2. This found roughly two-thirds of the stars in the data sets. If I spent more time on the cross matches I could probably find more.

This is very much a draft map. There are many possibilities for errors with this kind of map: errors in the ionizing stars for specific HII regions, bad cross matches, etc.

Also, a better map would be based on Gaia DR2 star density and not just a specific list of stars. And would be 3D. All of that is under development by myself and others. Expect to see some amazing Milky Way maps over the next few weeks and months!

Pages

Subscribe to RSS - Kevin Jardine's blog