What is a Galaxy?

“Any of the numerous large groups of stars and other matter that exist in space as independent systems.” (OED)

“A galaxy is a massive, gravitationally bound system that consists of stars and stellar remnants, an interstellar medium of gas and dust, and an important but poorly understood component tentatively dubbed dark matter.” (Wikipedia)

“I’ll know one when I see one” (Prof. Simon White, Unveiling the Mass of Galaxies, Canada, June 2009)

The question of “What is a galaxy” is being debated online at the moment, after it was posed by two astronomers – Duncan Forbes and Pavel Kroup in a paper posted on the arXiv last week. It’s an article written for professional astronomers, so doesn’t shirk the technical language in the suggestions for definitions, but in a very “zoo like” fashion (and following the model of the IAU vote on the definition of a planet which took place in 2006) invites the readers of the paper to vote on the definition of a galaxy. This has been reported in a few places (for example Science, New Scientist) and everyone is invited to get involved in the debate.

In fact Galaxy Zoo is cited in the press release about the work as one of the inspirations to bring this debate to a vote.

So what’s all the fuss about? Well it all started because of some very tiny galaxies which have been found in the last few years. There has been a debate raging in the scientific literature over whether or not they differ from star clusters, and where the line between large star clusters and small galaxies should be drawn. It used to be there was quite a separation between the properties of globular clusters (which are spherical collections of stars found orbiting galaxies – the Milky Way has a collection of about 150-160 of them) and the smallest known galaxies.

The globular cluster Omega Centauri. Credit: ESO

For example globular clusters all have sizes of a few parsecs (remember 1 parsec is about 3 light years), and the smallest known galaxies used to all have sizes of 100pc or larger. Then these things called ‘ultra compact dwarfs’ were found (in 1999), which as you might guess are dwarf galaxies which are very compact. They have sizes in the 10s of parsec range, getting pretty close to globular cluster scales.

UCDs in the Fornax Cluster. The background image was taken by Dr Michael Hilker of the University of Bonn using the 2.5-metre Du Pont telescope, part of the Las Campanas Observatory in Chile. The two boxes show close-ups of two UCD galaxies in the Hilker image. (Credit: These images were made using the Hubble Space Telescope by a team led by Professor Michael Drinkwater of the University of Queensland.)

Such objects begin to blur the line between star clusters and galaxies.

And there are things which have been called ‘ultra-faint dwarf spheroidal (dSph) galaxies’. These look nothing like the kind of galaxies you’re used to in SDSS images, although they were found in SDSS data – but perhaps not how you might expect. Researchers colour coded the stars in SDSS by their distance, and looked for overdensities or clumps of stars. So far several concentrations of stars at the same distance have been found. Some were new star clusters, but some look a bit like galaxies. If these are galaxies they are the smallest know, with only 100s of stars. Some are so faint that they would be outshone by a single massive bright star.

A map of stars in the outer regions of the Milky Way Galaxy, derived from the SDSS images of the northern sky, shown in a Mercator-like projection. The color indicates the distance of the stars, while the intensity indicates the density of stars on the sky. Structures visible in this map include streams of stars torn from the Sagittarius dwarf galaxy, a smaller 'orphan' stream crossing the Sagittarius streams, the 'Monoceros Ring' that encircles the Milky Way disk, trails of stars being stripped from the globular cluster Palomar 5, and excesses of stars found towards the constellations Virgo and Hercules. Circles enclose new Milky Way companions discovered by the SDSS; two of these are faint globular star clusters, while the others are faint dwarf galaxies. Credit: V. Belokurov and the Sloan Digital Sky Survey.

The paper suggests as a minimum that a galaxy ought to be gravitationally bound, and contain stars. They point out that this definition includes star clusters as well as all galaxies, so suggest some additional criteria might be needed. They make a list of suggestions, which we are invited to vote on. For full details (and if you are technically minded) I refer you to the paper. It’s a lesson in gravitational physics in itself (although as I said it is aimed at professional astronomers). Here is my potted summary of the suggestions they make:

1. Relaxation time is longer than the age of the universe. Basically this means that the system ought to be in a state where the velocity of a star in orbit in it will not change due to the gravitational perturbations from the other stars in a “Hubble time” (astronomer speak for a time roughly as long as the age of the universe).  This will exclude star clusters which are compact enough to have shorter relaxation times, but makes UCDs and faint dSph be galaxies.

2. Size > 100 pc (300 light years). Pretty self explanatory. Sets a minimum limit on the size. This makes the UCDs not be galaxies.

3. Should have stars of different ages. (a “complex stellar population”). The stars in most star clusters are observed to have formed all in one go from one massive cloud of gas. But in more massive systems not all the gas can be turned into stars at one time, so the star formation is more spread out resulting in stars of different ages being present. However there are some (massive) globular clusters which are know to have stars of different ages, so they would become galaxies in this definition.

4. Has dark matter. Globular clusters show no evidence for dark matter (ie. their measured mass from watching how the stars move is the same as the mass estimated by counting stars), while all massive galaxies have clear evidence for dark matter. The problem here is that this is a tricky measurement to make for UCDs and many dSph, so will leave a lot of question marks, and may not be the most practical definition.

5. Hosts satellites. This suggests that all galaxies should have satellite systems. In the case of massive galaxies these are dwarf galaxies (for example the Magellanic clouds around the Milky Way), and many dwarf galaxies have globular clusters in them. But there are some dwarf galaxies with no known globular clusters, and UCDs and dSph do not have any.

The paper finished by describing some of the most uncertain objects and provides the below table to show which would be a galaxy under a given definition. I’ve tried to explain what some of these all are along the way, and here is a short summary:

Omega Cen (first image above) is traditionally thought to be a globular cluster (ie. not a galaxy). Segue 1 is one of the ultra faint dSphs found by counting stars in SDSS images (3rd image). Coma Berenices is a bit brighter than a typical ultra faint dSph, and a bit bigger than a typical UCD. VUCD7 is (the 7th) UCD in the Virgo cluster (not the most imaginative name there!), so simular to the Fornax cluster UCDs shown in the second image. M59cO is a big UCD – almost a normal dwarf galaxy but not quite. BooII (short for Bootes III) and VCC2062 are objects which may possibly be material which has been tidally stripped off another galaxy. Or they might be galaxies.

Anyway if you’re interested you can join in the debate here. And if you read the paper you can vote. I voted,  but I think in the interest of letting you make up your own mind I’m not going to tell you what my decision was.

Oh and I just started a forum topic on it in case you want to debate inside the Zoo before voting.