Blue Sky and Red Spirals
This post is from Karen Masters at Portsmouth, who is working on red spirals….
When light travels through stuff it is scattered and absorbed. This is true of light passing through our atmosphere, and it is also true of light as it passes through galaxies. Light of different wavelengths is affected by this scattering and absorption in different ways. Bluer (or shorter wavelength) light is easier to scatter. The sky is blue on a cloudless day because the bluer light from the Sun is scattered out of the line of sight. This light “bounces around” off atoms and molecules in our atmosphere and eventually reaches our eye from some random direction – making the sky look blue. Obviously the light from the Sun itself appears slightly reddened by the same effect since the blue light is preferentially removed. At sunset or sunrise, when the Sun is close to the horizon the light from the Sun has to take a longer path through the atmosphere to get to us. More scattering takes place and the Sun appears redder than normal and makes a beautiful sight to see.
In spiral galaxies, the length of the path the light takes through the galaxy before it gets out and heads towards us depends on our viewing angle. When we see a spiral galaxy face-on the light has the shortest possible path out of the galaxy. By contrast, in an edge-on galaxy, the light must travel through most of the disk before getting out. We expect then that if scattering is important, edge-on galaxies will appear to be redder than face-on galaxies – for similar reasons that sunsets are red. The big question here though is “is scattering important”. Put another way we want to ask “are the disks of spiral galaxies transparent?”. We enjoy a fairly clear view of the extragalactic sky out of our spiral galaxy (the Milky Way), which suggested to early researchers than spiral galaxies probably were transparent. However it is also clear that there is a lot of “extinction due to dust” (our Astronomers terminology for the effect of scattering and absorption of light by particles in the inter-stellar medium) when we look towards the Galactic centre.
So what’s the problem of just looking at a bunch of spiral galaxies and seeing if they get redder as they get edge-on? Well nothing… except that you need to know you’re definitely looking at spirals, and you need to figure out how to measure how edge-on the spirals are. This of course is where Galaxy Zoo helps out so much. Thanks to you we now have an enormous number of visually classified unquestionably spiral galaxies. You even picked out the edge-on ones for us. We can also use the “axial ratios” (the ratio of the maximum dimension to the minimum dimension) of the galaxies from Sloan, which (with some assumptions about how thin the average galaxy is when it’s totally edge-on) gives an estimate of the exact angle of the galaxy’s orientation to us.
And what we’re finding is that spirals definitely get redder as they get more edge-on. So extinction due to dust is clearly important. Because Sloan measures the galaxies in 5 different wavelengths, we can make 4 Sloan colours (in Astronomy the colour is just the difference in the brightness in two different bands) and look at the relative amount of extinction with wavelength which provides information on the source of the scattering and absorption. We can also go to other surveys (for example UKIDSS which measures near-infra red light) to extend this further for some of the galaxies.
Extinction seems to be quite a hot topic lately with Sloan data, but what we have which other researchers don’t is the Galaxy Zoo classifications. They have to use other estimates of if the galaxy is a spiral or not, such as how concentrated the light is, or the exact details of the light profile. Neither is as simple or as reliable as having a human just look at the galaxy. Measuring the amount of extinction is important because it’s been largely neglected in studies using Sloan data up until now. The physical parameters of a galaxy ought not to depend on our viewing angle, but when researchers use colours and luminosities to estimate the star formation history or stellar mass of a spiral galaxy the answer will depend on viewing angle if extinction due to dust is not corrected for. More importantly, elliptical galaxies do not suffer from this effect, so if you compare the mean properties of ellipticals and spirals your answer will be biased by the effect of dust.
So most red spirals seem to be edge-on dusty star forming galaxies which would be normal blue spirals if seen face-on…. but this can’t explain all red spirals. We can still see a significant population of red face-on spirals, and by measuring the average amount of reddening we will even be able to pick out the edge-on spirals which would still be red if seen face-on.
I moved to Portsmouth in October and I was delighted to start working with the Galaxy Zoo team and data. I knew about the project (and even classified a handful of galaxies) before I moved here. I’m currently working on a short paper describing what I’ve told you about here and hope to have it submitted early in the New Year.
Merry Christmas!
Some example images:
A blue face-on spiral galaxy.
A red face-on spiral galaxy.
A red edge-on spiral galaxy.
10 responses to “Blue Sky and Red Spirals”
Trackbacks / Pingbacks
- - November 12, 2009
Galaxy Zoo 
Karen, thank you for all of this information, which is wonderfully engaging and clear, but I’m flummoxed by your example images.
I would say that face-on galaxy number two is less blue than face-on galaxy number one, but I don’t see anything in it that I would describe as red.
There are, in the GZ red spirals thread, dozens of spiral galaxies that are unmistakably red and not blue, so I wonder why you chose the one you did for example number two. But perhaps there is something about your use of “red” that I don’t understand?
Karen what wonderful information, I am looking forward to reading more about this subject.
Hi LankyYankee,
I’m not aware exactly how Karen splits her spirals into red and blue, but broadly speaking it’s worth remembering that SDSS has five filters and therefore four colours (we need to know the relative brightness in each, so of u, g, r, i and z filters the ‘colours’ are u-g, g-r, r-i and i-z). You’d usually look at how things were spread out in each of these and find some sensible way to split your group up – for example if you have two definite clumps of colour where objects tend to sit you’d divide between those two. Then the redder of these you’d call ‘red’ and the bluer you’d call ‘blue’.
So these ‘red’ and ‘blue’ are relative terms, and might not translate directly into very different looking images on the screen. So you might well call these spirals ‘less blue’.
Remember the SDSS images aren’t a direct representation of what the human eye would see either.
Thanks Karen and good luck with the paper. 🙂
Thanks for all the nice comments, and thanks to Edd for responding to LankyYankee by saying exactly what I would have! By red spirals here I’m using a definition which says that their SDSS u-r colours are redder than a line defined in a study of the colours of SDSS galaxies published in 2006 (Baldry etal 2006; MNRAS 373, 469 for those who might care!). To find examples of red and blue face-on spirals to post here I looked for very round galaxies which were quite far from that line in both directions, but to our eyes I admit the difference is quite subtle. Also bear in mind I’m using the “integrated colours” (ie. averaged over the whole galaxy). In the images above my eye is drawn to the fact that the blue face-on spiral looks quite “red” in the centre (as is expected for a spiral), but if you look closely its spiral arms are bluer (albeit subtly bluer to the eye) than the arms in the red face-on spiral.
Ooops, during my first days of classification, I put some of the “fried egg” galaxies into edge-on/unclear on the basis that they were unclear. I’m sure this was too late to affect your paper, but can I apologize in advance for affecting the next paper’s correlation coefficient.
Budgieye – I’m not an expert on the classifcation, but I wouldn’t worry. It’s my understanding that each galaxy is classified by many people on the site and then if there are any odd classifications (like what you describe) they’d probably get tossed out. That’s the beauty of having so many people interested in classifying.
This wouldn’t affect my work anyway as I use the “roundness” of the galaxy (ie. how close it is to circular) to estimate the inclination. A face-on galaxy will be very round of course, while an edge-on one will be very far from round.
Hi Karen,
just started Zoo 2, browsing the forums. I understand the scattering of blue light leaves the sun looking red at sunset and sunrise – why does the moon not look red when setting and rising? Especially at night, when there is no sunlight (direct or scattered) to drown out any colour change?
an explanation would be most welcome!
Hi Bob. Sorry it took so long to notice your comment here. The Moon does sometimes look red (or orange) when it’s close to the horizon, however it’s not as easy for our eyes to distinguish colour information at low light levels, so colour appears much weaker. That’s why you might not have noticed any strong colour change in the Moon.