What is a galaxy cluster

1
What is a galaxy cluster?
Aaron Fenyes (REU 2007)Advisor Tim McKay
Reality
Theory
2
(No Transcript)
3
(No Transcript)
4
(No Transcript)
5
Three ways to describe a cluster

• A group of galaxies
• A cloud of hot gas
• A clump of dark matter

6
Galaxy-based cluster finding

• Advantages galaxies are easier to see than gas,
  and there dont seem to be any clusters without
  galaxies in them
• The bad news you will be forever haunted by
  projection effects
• The good news projection effects can be overcome
• Matched filters
• Cluster red sequence

7
Matched filters

• Introduced by Marc Postman in 1996
• The filter is a model of what a galaxy cluster
  at a given redshift ought to look like
• Angular position distribution
• Magnitude (brightness) distribution
• By comparing the model to various patches of sky,
  you can build a likelihood map
• By building likelihood maps for several different
  model redshifts, you can distinguish between
  nearby clusters, faraway clusters, and chance
  projections that arent really clusters at all
• Bonus you get a redshift estimate for each
  cluster!

8
Cluster red sequence

• Introduced by Michael Gladders and H. K. C. Yee
  in 2000
• Exploits the fact that early-type galaxies have
  a very predictable color-magnitude relationship
• Star formation ends almost immediately after the
  galaxy comes together
• Ram pressure stripping?
• Beats projection by looking for galaxies that are
  clustered in both angular position and color
• Bonus you get a redshift estimate for each
  cluster!

Redder
Brighter
9
Three ways to describe a cluster

• A group of galaxies
• A cloud of hot gas
• A clump of dark matter

10
The halo model

• The equations that govern the expansion and of
  the entire universe also apply to any spherical
  region within the universe
• In a flat universe, overdense regions will
  recollapse while the rest of the universe keeps
  expanding around them
• After a while, most of the mass in the universe
  will have collapsed into small, spherical halos
  of dark matter

11
N-body simulations

• Each simulation particle represents a large blob
  of dark matter
• How do we connect the blobs?
• Spherical overdensity
• Voronoi tessellation
• Friends-of-friends

12
Back to the point testing the cluster finder

• How do we know that the galaxy groups were
  finding are really clusters?
• From a theorists point of view, a real cluster
  is a dark matter halo
• By running the cluster finder on simulated data,
  we can see how many of the groups we find
  correspond to halos

13
Back to the point testing the cluster finder

• The catch n-body simulations dont contain any
  galaxies!
• The solution pick a sample of dark matter blobs
  and attach fake galaxies to them
• We have to be careful, because the distribution
  of galaxies in the real universe doesnt follow
  the distribution of dark matter exactly
• We also have to make sure that our fake galaxies
  look like real cluster galaxies, with the right
  colors and magnitudes
• Most real clusters have a big, bright galaxy
  sitting at the center, and we have to reproduce
  this too
• This is done by Risa Wechsler, now at Stanford
  University

14
And I get the easy job

• Eduardo Rozo, at Ohio State, has come up with a
  method for evaluating cluster finder performance
  based on the simulated runs we just talked about
• Used it to evaluate the catalog constructed by
  Ben Koester, a previous grad student in our lab
• All I had to do was re-implement his method and
  run it on preliminary versions of Jiangangs
  catalog

15
Some typical results
16
Acknowledgments

• Many thanks to Jiangang Hao, Tim McKay, Matt
  Becker, Brian Nord, Eli Rykoff, Paige Warmker and
  the rest of the cluster crew (in order of the
  amount they had to put up with me)