Background Information

Colors

Astronomers like to say that a galaxy is "blue" or that a star is "red". In some cases we are actually saying "that star looks red", but most of the time we are referring to the object's color. The color of a galaxy, for instance, is not necessarily related to its visible appearance. When we say the sky is blue, we mean that the sky looks blue. When we say a galaxy is blue, we are referring to the difference in magnitude between two different filters. For example, if a galaxy has a magnitude in the SDSS g' filter of 8.42 and magnitude in the i' filter of 12.56, then the g-i color is -4.14. Technically, color is more than just the difference of two magnitudes; it is actually a ratio of the flux of an object as measured in two filters. Magnitude itself is a function of an object's flux, and when we subtract two magnitudes, we are really dividing by their fluxes, since magnitude is logarithmic. For example, we can write the magnitude difference of two stars as:

m1 - m2 = -2.5log(F1/F2)

We can substitute filter magnitudes in for star magnitudes, and we get the same relationship.

Now back to our original discussion: what makes a galaxy "blue"? Well, that all depends on the color of the galaxy. Our example galaxy has a g-i color of -4.14. Because the color is negative, we know there is more light coming from the g' filter than the r' filter. If we look at a graph showing the wavelength coverage of each filter, we can see that the g' filter covers wavelengths towards the blue end of the spectrum while the i' filter covers wavelengths at the red end of the spectrum. Therefore, this is a "blue" galaxy. If instead the galaxy had a color of +4.14, we would say it was a "red" galaxy. You may be wondering about the "yellow" or "green" or "magenta" galaxies, but we only use "blue" or "red". There is no definite color value where we can classify an object as red or blue, especially since we may not use the same two filters every time, but we can generally call any object with a color of ~1 or less blue. For the u-r color, there is a nicely defined cutoff for blue and red galaxies at u-r = 2.2, but it is not so nice for others.