One of the joys, frustrations, most loved, and most hated parts of being a professor is attempting to do research. I say attempting because sometimes the data just doesn’t want to produce anything useful.
There are good times. For instance, in about three months this summer and fall Fraser Cain and I, with the help of undergraduate Rebecca Bemrose-Fetter and graduate student Georgia Bracey, managed to do a quick a solid study on who listens to Astronomy Cast and responds to surveys. The paper is already published and you can find it here. That was fun, challenging to analyze in a “I need a brain but not a Nobel prize” kind of way. That’s the type of low-hanging-fruit every researcher likes to pick and munch every now and again.
The research I worked on today, however, led me to question the logical nature of my field. Along with another student (whose name I’m not publishing without permission), I’m currently working on a follow-up project to my dissertation. Back in 2002, I noticed (and I’m not the only one to notice this) that there is a trend as a function of density in how galaxy clusters evolve. Basically, small groups of galaxies, like the one we live in, don’t really evolve that much over the course of the universe. Spiral galaxies stay spiral and violence like collisions and gravitational harassment just don’t happen that often. At the same time, really rich large clusters form first and very quickly and they whip their member galaxies into a frenzy of mutual destruction. Galaxies are quickly beaten into elliptical forms, star formation is cut off, and anything new that falls in is quickly destroyed. It is the mid-sized systems that are most interesting. They start out filled with spirals, but as the millenniums tick by the spirals collide into one another, one by one, until today these systems are devoid of star formation and rich boring elliptical galaxies.
Observationally, we have specific ways to describe cluster density and the fraction of spiral to elliptical galaxies. Unfortunately, we have more than one way to do each of these things. This is where my personal frustration comes in. My student and I wanted to cull for the astronomy literature a large collection of published values and than look for a general trend. We were even prepared to solve for ways to convert from one way of determining galaxy density to another way. What we hadn’t expected is the utter lack of relationship we are finding in some cases. Let me see if I can explain at least one aspect of this problem.
Let’s take galaxy size and density. The most well known catalogue of galaxy clusters is probably the Abell Catalogue. In his original 1958 paper, Abell described cluster richness using the number of galaxies “counted in a cluster that are not more than 2 mag. fainter than the third brightest member.” This method requires the counter to know where the edge of the galaxy is, and doesn’t work very well for large, diffuse systems that are difficult to sort out from background and foreground systems (think, Zwicky clusters). It also doesn’t work well for systems that are rich, but only have a few extraordinarily bright galaxies and scores of fainter systems.
Since that catalogue, people have been trying over and over to find a better way. For instance, Butcher and Oemler count galaxies “with projected distances from the cluster center less than R30, and with absolute visual magnitudes Mv <= -20” where R30 “is the radius of the circle containing [30]% of the cluster’s projected galaxy distribution.” This system consistently counts the same type of galaxies, and while R30 isn’t perfect, it is better than trying to define the whole galaxy cluster.
Other methods also measure the number of galaxies brighter than certain cutoff magnitudes within a specific number of megaparsecs, or the number of galaxies within a specific number of megaparsecs of the cluster’s brightest galaxy or radio galaxy.
I’ve come to the conclusion that there are as many ways of measuring cluster size as there are groups measuring cluster parameters. The radical differences in the systems means there isn’t even a direct way to convert from one system to another. This means that it just isn’t possible to look for detailed relationships using all the data that is out there without doing some serious reanalysis. What had looked like a nice easy literature review project to work on with a student grew into something that is requiring a lot of head scratching, and I’m afraid that to do the project right, we’ll need to use SDSS data, which makes this very much not a 1 semester project for an undergrad. So… we’re going to do what we can with all the published data that we can get on a mostly standard system (and it looks like we’ll be using Abell counts, as flawed as they are).
At least it’s not quite apples and oranges. I think I can safely call all our data citrus. Unfortunately, I think the clemintines got mixed in with the tangerines.
Image Credit: That ones all mine 🙂 It’s a galaxy cluster I discovered as part of my dissertation research.
I know research can be frustrating but I never realized it could be *that* frustrating.
It is always interesting when professionals take the time to illuminate the potholes along the adventurous road. Thanks.
But potholes are part of science, too. Perhaps the lack of correlation you’ve discovered is a worthy finding in itself.
You discovered a GALAXY CLUSTER??!?!?!?!?
And??? And??? Where is it? What is its designation???
You discover a GALAXY CLUSTER and you don’t tell me??? 🙂