The color of galaxies, and the stars that make them
Posted: Sat May 07, 2011 6:43 pm
So I'm going to talk a bit more about my favorite subject, the colors of galaxies.
Galaxies are, of course, made up of all kinds of things. All galaxies, including ellipticals, contain gas and dust, but in the case of the ellipticals the dust generally doesn't form visible dust clouds. All galaxies that contain visible dust clouds have clouds that are dark, but galaxies that form stars also have clouds that shine by reflecting light from bright stars, and galaxies with star formation also have gas clouds that are situated so close to newborn hot star that they get their electrons knocked about to the point that they shine by their own light.
We can be sure that all galaxies contain large collections of planets, moons, asteroids, comets and all kinds of "debris".
And then there are the stars, of course. There are itty bitty stars that hardly shine at all, the small red dwarfs and the even smaller brown dwarfs. There are the even smaller white dwarfs, which shine fairly brightly when they are young, but then just gradually fade and decline, however slowly. And there are stars which have exploded brilliantly and formed really tiny neutron stars, and there are black holes, some of which are "quiet" and invisible, some of which are supergiant and violent, prone to violent outbursts and surrounded by brilliantly luminous accretion disks.
And there are the bright stars, the stars that are mainly responsible for making galaxies shine in the night. What stars are they?
I am the owner of a a book called Sky Catalogue 2000.0, which lists 50,071 stars. Of these, the largest group is K-type stars. 14,534 are mentioned. The second largest group is A-type stars, 8,189 individuals. Next come the F-type stars: 7,321. Then we have the G-type stars: 5,847. Then the B-type ones: 4,116. Then the M-type ones: 1,835. Finally the O-type stars: only 125 listed.
Of course, if you summarize these numbers you get only 41,967 stars. But that is because the number of stars belonging to the various spectral classes is presented as a bar chart, going from stars of apparent magnitude zero (Sirius) down to stars of apparent magnitude eight, and the bars showing the numbers of stars of the brightest magnitude slots are so short that no figure was given for the number of stars belonging to them.
http://iopscience.iop.org/1538-3881/125 ... 8.fg2.html
I found this histogram on the net, showing the number of stars belonging to different spectral classes and luminosity classes. This histogram differs from the bar chart in Sky Catalogue 2000.0 in that there are so many stars belonging to spectral class G and luminosity class V, but otherwise these two histograms are similar. Note that luminosity class III generally generally means a greater luminosity than luminosity class V, so that the K-type stars of luminosity class III will be much, much brighter than the G-type stars of luminosity class V.
An interesting fact can be concluded from these numbers. In a large galaxy where star formation has continued uninterrupted perhaps since the first building blocks of our galaxy were put together, and where there is an ever increasing metal enrichment of the available star forming material, there is going to be a very, very large population of K-type, helium-fusing red giants, which are going to contribute an awful lot to the overall light of the galaxy. There is, of course, going to be an even larger population of K-type dwarfs, but because they are so intrinsically faint they will probably not contribute very much to the overall light of the galaxy.
The typical bright K-type giant is a star that has ceased hydrogen fusion in its center and started fusing helium in its core instead. These stars form wherever you have a moderately massive star that has used up the hydrogen in its center and made itself hot enough to start fusing helium. Such stars form everywhere and not necessarily in clusters. Two factors contribute to the ubiquitousness of these K-type stars. First, all stars that are sufficiently massive will once become K-type giants, and the Sun, too, will one day become such a star. Second, if you have any star formation at all in a galaxy, then this star formation will always produce stars that are sufficiently massive to eventually turn into K-type giants. So the production of K-type giants is an ongoing affair in basically all galaxies in the known universe. This production of K-type giants will go on in every galaxy until the last star that is sufficiently massive to use up the hydrogen in its core and get its helium fusion going has used up the helium in its core and evolved further.
The typical A-type star in a galaxy, by contrast, is going to be a hydrogen fusing dwarf. The A-type dwarf needs to be more massive than the K-type giant has to be. The Sun has never come close to becoming an A-type dwarf, but it will have no trouble turning itself into a K-type giant. Therefore the A-type dwarfs are less common than the K-type giants. Also the A-type dwarfs are more "localized" than the K-type giants, and you often find them in clusters. But since the A-type dwarfs are not very massive, they are not necessarily strongly clustered. A-type dwarfs don't live for more than, say, 500 million years of so. They don't survive for as long as a billion years. Therefore, if a galaxy has had no star formation for a billion years, it will not contain any A-type stars, unless we are talking about true oddities like the so-called "blue stragglers".
M11, a bright cluster dominated by A-type stars. Credit: Chris Deforeit.
F- and G-type dwarfs are more common and live longer than the A-type dwarfs, but they are fainter than the A-type dwarfs and don't contribute very much of the overall light of a galaxy, unless they exist in extremely large numbers. As for the K-type main sequence stars, I'd say they are too faint to make a contribution to the overall light of a galaxy, and this is even more true of the extremely numerous but very, very faint M-type dwarfs.
Credit: ESO.
The arrow points at Proxima Centauri, a typical faint M-type dwarf. These stars don't make much of a bang. The bright stars at upper left are Alpha Centauri A and Alpha Centauri B. (Alpha A is a G-type dwarf and Alpha B is a K-type dwarf, by the way.) Proxima is a little closer to us than than Alpha Centauri A and B.
What about M-type giants and supergiants, though? Yes, they exist and they can be supremely luminous, but there are so few of them that they will never affect the overall color of a galaxy.
What about the rare OB stars? There are far more OB supergiants than M-type supergiants, and the OB stars will affect the color of a galaxy a lot, if they exist in even moderate numbers, because they are so bright.
Credit: Matthew Spinelli.
Orion contains only one red M-type supergiant compared with at least seven hot blue O- and B-type giants and supergiants. There are also large numbers of main sequence O- and B-type stars, including Theta 1C Orionis, the star powering the Orion Nebula. Therefore and overall, Orion is a "blue" and not a "red" constellation.
OB stars are virtually always born in clusters, and at least the O-type stars are sufficiently short-lived that they are going to stay in clusters. Therefore O-type stars are never evenly mixed in a galaxy, but instead they are always localized into clusters, giving a galaxy with a lot of O-type stars a clumpy appearance.
Credit: ING Archive and Nik Szymanek.
The knotty and clumpy appearance of NGC 4038 demonstrates that this galaxy is home to many large and brilliant clusters and many highly luminous O-type stars. NGC 4039, on the other hand, contains far fewer clusters, and those that exist are generally not so bright. There are certainly not as many O-type stars in NGC 4039 as there are in NGC 4038.
The color and light distribution of a galaxy depends, apart from the presence of dust and nebulae, on the type of stars the galaxy contains, the relative number of them and how they are mixed.
The reddest galaxies in existence, apart from some very strongly dust-enshrouded Ultra Luminous Infrared Galaxies, are the elliptical galaxies, because they contain virtually no O, B and A-type stars at all. Most of their light comes from K-type giants. The overall color of elliptical galaxies are typically not quite as yellow as K-type giants, but a little bit "whiter". My guess is that the yellow color of the K-type giants has been somewhat diluted by the presence of significant numbers of G-type dwarfs like the Sun.
M86, a giant elliptical galaxy whose light comes mainly from K-type giant stars. Credit: Sloan Digital Sky Survey.
Shouldn't the color of M86 be yellower? Probably not. Its B-V index is 0.93 and its U-B index is 0.65. Compare it with K0 star Pollux, whose B-V index is 0.99. Mix stars like Pollux with some G-type stars like the Sun, and you are likely to get a color similar to that of galaxy M86.
Constellation Gemini with yellow-orange K0 star Pollux farthest to the left. Credit: JOHN SANFORD/SCIENCE PHOTO LIBRARY.
Of course, there are dwarf galaxies whose population is extremely old, but which are noticeably bluer than the large ellipticals. That is because some of the dwarf lenticulars (as they are often called) may be extremely metal-poor, and if so many of their K-type giants will turn into blue horizontal stars instead.
Credit: Hubble Space Telescope.
This is a picture of Omega Centauri, the brightest globular cluster of the Milky Way. Astronomers suspect that Omega Centauri may have been the bulge of an ancient dwarf galaxy.
Adam Block recently posted a brilliant image of galaxies NGC 5614 and 5613 here at Starship Asterisk:
Here is a Sloan Digital Sky Survey image of the same galaxy pair:
This pair of galaxies are among the few I can think of where an F- and G-type population likely make a noticeable and localized contribution to the overall light of the galaxy. Note in the larger galaxy a bluish, but not very blue, ring, whose light may well be strongly influenced by F-type stars. In the smaller galaxy at top, there is a ring surrounding the nucleus. This ring is sufficiently neutral in color that it may well be mostly influenced by G-type stars.
The reason why I'm guessing that NGC 5614 may contain a large F-type population in its ring is not just because I think that this can be concluded from the Sloan Digital Sky Survey image above and from the image by Adam Block. A very important reason for my assessment in this case is the UBV image of NGC 5614 in James D. Wray's Color Image of Galaxies. Wray took very great care to calibrate the color of his pictures, both by photographing many of the galaxies with two different telescopes and by comparing the color of his galaxies with the color of moderately bright foreground stars, whose color indexes could be determined. That's why I trust the color balance of his pictures a lot.
Wray's picture of NGC 5614 shows the ring surrounding the yellow bulge to be yellow-green in color in UBV. That makes it virtually certain that extremely few if any OB stars are present in the ring. The presence of a significant population of A-type stars can't be ruled out, but if so, these A-type stars must be very evenly mixed with an unusually large population of K stars to produce the yellow-green color in UBV. I find that less likely than the idea that the K-type stars are present in "normal" numbers and are mixed primarily with large numbers of F-type stars, the remnants of a great starburst perhaps a billion years ago. As for my guess that the ring surrounding the nucleus in NGC 5613 contains a large number of G-type stars, that assumption is based solely on the fact that the ring of NGC5613 looks yellower than the ring of NGC 5614, but the ring of NGC 5613 is still not as yellow as the inner bulge of this galaxy, certainly not in Adam Block's image.
A rather fantastic example of a galaxy that may contain a huge number of A-type stars, but few OB stars, is NGC 5394.
NGC 5394 by SDSS
Note the incredible blue arc to the left of the bulge in NGC 5394, the small galaxy at top. The arc is really bright, but apparently completely smooth. The distribution of O-type stars is never smooth, and therefore the blue arc probably contains no O stars at all. B stars may live long enough to get a lot more mixed, so the arc might contain B-type stars. However, in James D. Wray's Color Atlas of Galaxies the arc of NGC 5394 is all green in color - in fact, it is even slightly yellow-green - which suggests that it is made up of A- as well as F-type stars, mixed (as always) with K-type giants.
In my next post here I will say something about how the colors of galaxies are shown in images through the use of different filters. I will also discuss why filters are sometimes used which seem to "shift" the color of the galaxy away from what we would see when we looked at at, if our color vision was the same as it is now but dozens or hundreds of times more sensitive.
Ann
Galaxies are, of course, made up of all kinds of things. All galaxies, including ellipticals, contain gas and dust, but in the case of the ellipticals the dust generally doesn't form visible dust clouds. All galaxies that contain visible dust clouds have clouds that are dark, but galaxies that form stars also have clouds that shine by reflecting light from bright stars, and galaxies with star formation also have gas clouds that are situated so close to newborn hot star that they get their electrons knocked about to the point that they shine by their own light.
We can be sure that all galaxies contain large collections of planets, moons, asteroids, comets and all kinds of "debris".
And then there are the stars, of course. There are itty bitty stars that hardly shine at all, the small red dwarfs and the even smaller brown dwarfs. There are the even smaller white dwarfs, which shine fairly brightly when they are young, but then just gradually fade and decline, however slowly. And there are stars which have exploded brilliantly and formed really tiny neutron stars, and there are black holes, some of which are "quiet" and invisible, some of which are supergiant and violent, prone to violent outbursts and surrounded by brilliantly luminous accretion disks.
And there are the bright stars, the stars that are mainly responsible for making galaxies shine in the night. What stars are they?
I am the owner of a a book called Sky Catalogue 2000.0, which lists 50,071 stars. Of these, the largest group is K-type stars. 14,534 are mentioned. The second largest group is A-type stars, 8,189 individuals. Next come the F-type stars: 7,321. Then we have the G-type stars: 5,847. Then the B-type ones: 4,116. Then the M-type ones: 1,835. Finally the O-type stars: only 125 listed.
Of course, if you summarize these numbers you get only 41,967 stars. But that is because the number of stars belonging to the various spectral classes is presented as a bar chart, going from stars of apparent magnitude zero (Sirius) down to stars of apparent magnitude eight, and the bars showing the numbers of stars of the brightest magnitude slots are so short that no figure was given for the number of stars belonging to them.
http://iopscience.iop.org/1538-3881/125 ... 8.fg2.html
I found this histogram on the net, showing the number of stars belonging to different spectral classes and luminosity classes. This histogram differs from the bar chart in Sky Catalogue 2000.0 in that there are so many stars belonging to spectral class G and luminosity class V, but otherwise these two histograms are similar. Note that luminosity class III generally generally means a greater luminosity than luminosity class V, so that the K-type stars of luminosity class III will be much, much brighter than the G-type stars of luminosity class V.
An interesting fact can be concluded from these numbers. In a large galaxy where star formation has continued uninterrupted perhaps since the first building blocks of our galaxy were put together, and where there is an ever increasing metal enrichment of the available star forming material, there is going to be a very, very large population of K-type, helium-fusing red giants, which are going to contribute an awful lot to the overall light of the galaxy. There is, of course, going to be an even larger population of K-type dwarfs, but because they are so intrinsically faint they will probably not contribute very much to the overall light of the galaxy.
The typical bright K-type giant is a star that has ceased hydrogen fusion in its center and started fusing helium in its core instead. These stars form wherever you have a moderately massive star that has used up the hydrogen in its center and made itself hot enough to start fusing helium. Such stars form everywhere and not necessarily in clusters. Two factors contribute to the ubiquitousness of these K-type stars. First, all stars that are sufficiently massive will once become K-type giants, and the Sun, too, will one day become such a star. Second, if you have any star formation at all in a galaxy, then this star formation will always produce stars that are sufficiently massive to eventually turn into K-type giants. So the production of K-type giants is an ongoing affair in basically all galaxies in the known universe. This production of K-type giants will go on in every galaxy until the last star that is sufficiently massive to use up the hydrogen in its core and get its helium fusion going has used up the helium in its core and evolved further.
The typical A-type star in a galaxy, by contrast, is going to be a hydrogen fusing dwarf. The A-type dwarf needs to be more massive than the K-type giant has to be. The Sun has never come close to becoming an A-type dwarf, but it will have no trouble turning itself into a K-type giant. Therefore the A-type dwarfs are less common than the K-type giants. Also the A-type dwarfs are more "localized" than the K-type giants, and you often find them in clusters. But since the A-type dwarfs are not very massive, they are not necessarily strongly clustered. A-type dwarfs don't live for more than, say, 500 million years of so. They don't survive for as long as a billion years. Therefore, if a galaxy has had no star formation for a billion years, it will not contain any A-type stars, unless we are talking about true oddities like the so-called "blue stragglers".
M11, a bright cluster dominated by A-type stars. Credit: Chris Deforeit.
F- and G-type dwarfs are more common and live longer than the A-type dwarfs, but they are fainter than the A-type dwarfs and don't contribute very much of the overall light of a galaxy, unless they exist in extremely large numbers. As for the K-type main sequence stars, I'd say they are too faint to make a contribution to the overall light of a galaxy, and this is even more true of the extremely numerous but very, very faint M-type dwarfs.
Credit: ESO.
The arrow points at Proxima Centauri, a typical faint M-type dwarf. These stars don't make much of a bang. The bright stars at upper left are Alpha Centauri A and Alpha Centauri B. (Alpha A is a G-type dwarf and Alpha B is a K-type dwarf, by the way.) Proxima is a little closer to us than than Alpha Centauri A and B.
What about M-type giants and supergiants, though? Yes, they exist and they can be supremely luminous, but there are so few of them that they will never affect the overall color of a galaxy.
What about the rare OB stars? There are far more OB supergiants than M-type supergiants, and the OB stars will affect the color of a galaxy a lot, if they exist in even moderate numbers, because they are so bright.
Credit: Matthew Spinelli.
Orion contains only one red M-type supergiant compared with at least seven hot blue O- and B-type giants and supergiants. There are also large numbers of main sequence O- and B-type stars, including Theta 1C Orionis, the star powering the Orion Nebula. Therefore and overall, Orion is a "blue" and not a "red" constellation.
OB stars are virtually always born in clusters, and at least the O-type stars are sufficiently short-lived that they are going to stay in clusters. Therefore O-type stars are never evenly mixed in a galaxy, but instead they are always localized into clusters, giving a galaxy with a lot of O-type stars a clumpy appearance.
Credit: ING Archive and Nik Szymanek.
The knotty and clumpy appearance of NGC 4038 demonstrates that this galaxy is home to many large and brilliant clusters and many highly luminous O-type stars. NGC 4039, on the other hand, contains far fewer clusters, and those that exist are generally not so bright. There are certainly not as many O-type stars in NGC 4039 as there are in NGC 4038.
The color and light distribution of a galaxy depends, apart from the presence of dust and nebulae, on the type of stars the galaxy contains, the relative number of them and how they are mixed.
The reddest galaxies in existence, apart from some very strongly dust-enshrouded Ultra Luminous Infrared Galaxies, are the elliptical galaxies, because they contain virtually no O, B and A-type stars at all. Most of their light comes from K-type giants. The overall color of elliptical galaxies are typically not quite as yellow as K-type giants, but a little bit "whiter". My guess is that the yellow color of the K-type giants has been somewhat diluted by the presence of significant numbers of G-type dwarfs like the Sun.
M86, a giant elliptical galaxy whose light comes mainly from K-type giant stars. Credit: Sloan Digital Sky Survey.
Shouldn't the color of M86 be yellower? Probably not. Its B-V index is 0.93 and its U-B index is 0.65. Compare it with K0 star Pollux, whose B-V index is 0.99. Mix stars like Pollux with some G-type stars like the Sun, and you are likely to get a color similar to that of galaxy M86.
Constellation Gemini with yellow-orange K0 star Pollux farthest to the left. Credit: JOHN SANFORD/SCIENCE PHOTO LIBRARY.
Of course, there are dwarf galaxies whose population is extremely old, but which are noticeably bluer than the large ellipticals. That is because some of the dwarf lenticulars (as they are often called) may be extremely metal-poor, and if so many of their K-type giants will turn into blue horizontal stars instead.
Credit: Hubble Space Telescope.
This is a picture of Omega Centauri, the brightest globular cluster of the Milky Way. Astronomers suspect that Omega Centauri may have been the bulge of an ancient dwarf galaxy.
Adam Block recently posted a brilliant image of galaxies NGC 5614 and 5613 here at Starship Asterisk:
Here is a Sloan Digital Sky Survey image of the same galaxy pair:
This pair of galaxies are among the few I can think of where an F- and G-type population likely make a noticeable and localized contribution to the overall light of the galaxy. Note in the larger galaxy a bluish, but not very blue, ring, whose light may well be strongly influenced by F-type stars. In the smaller galaxy at top, there is a ring surrounding the nucleus. This ring is sufficiently neutral in color that it may well be mostly influenced by G-type stars.
The reason why I'm guessing that NGC 5614 may contain a large F-type population in its ring is not just because I think that this can be concluded from the Sloan Digital Sky Survey image above and from the image by Adam Block. A very important reason for my assessment in this case is the UBV image of NGC 5614 in James D. Wray's Color Image of Galaxies. Wray took very great care to calibrate the color of his pictures, both by photographing many of the galaxies with two different telescopes and by comparing the color of his galaxies with the color of moderately bright foreground stars, whose color indexes could be determined. That's why I trust the color balance of his pictures a lot.
Wray's picture of NGC 5614 shows the ring surrounding the yellow bulge to be yellow-green in color in UBV. That makes it virtually certain that extremely few if any OB stars are present in the ring. The presence of a significant population of A-type stars can't be ruled out, but if so, these A-type stars must be very evenly mixed with an unusually large population of K stars to produce the yellow-green color in UBV. I find that less likely than the idea that the K-type stars are present in "normal" numbers and are mixed primarily with large numbers of F-type stars, the remnants of a great starburst perhaps a billion years ago. As for my guess that the ring surrounding the nucleus in NGC 5613 contains a large number of G-type stars, that assumption is based solely on the fact that the ring of NGC5613 looks yellower than the ring of NGC 5614, but the ring of NGC 5613 is still not as yellow as the inner bulge of this galaxy, certainly not in Adam Block's image.
A rather fantastic example of a galaxy that may contain a huge number of A-type stars, but few OB stars, is NGC 5394.
NGC 5394 by SDSS
Note the incredible blue arc to the left of the bulge in NGC 5394, the small galaxy at top. The arc is really bright, but apparently completely smooth. The distribution of O-type stars is never smooth, and therefore the blue arc probably contains no O stars at all. B stars may live long enough to get a lot more mixed, so the arc might contain B-type stars. However, in James D. Wray's Color Atlas of Galaxies the arc of NGC 5394 is all green in color - in fact, it is even slightly yellow-green - which suggests that it is made up of A- as well as F-type stars, mixed (as always) with K-type giants.
In my next post here I will say something about how the colors of galaxies are shown in images through the use of different filters. I will also discuss why filters are sometimes used which seem to "shift" the color of the galaxy away from what we would see when we looked at at, if our color vision was the same as it is now but dozens or hundreds of times more sensitive.
Ann