Finding AGN in cluster galaxies
For your thesis, we should look at the properties of AGN in
clusters. Why don't you go dig through the data archives and see
what you come up with?
Student: uh, OK.
So what do you do? This is an excellent example of a
multiwavelength data mining task. We'll combine optical and X-ray
data to study AGN in clusters.
Step I: Identify some
galaxy clusters. We'll start with the soft X-ray selected ROSAT
Brightest Cluster Sample, and I'm providing a easy to read data table. From that, we
want to select clusters that we also have optical and hard X-ray
Step II: Get the SDSS
optical photometry and spectroscopy of galaxies within 30 arcmin
radius of the cluster. Work out a good estimate for the cluster
redshift. Build a color-magnitude diagram for the cluster, first
for all (non-stellar) objects within 1 Mpc of the center, and then
overplot spectroscopically confirmed cluster members (you will
need to determine the cluster redshift and then also give your
criterion for "spectroscopically confirmed cluster member"). Use
these plots to identify the red sequence, and thereby assign
"photometric cluster membership".
Step III: Search the
Chandra source catalog for hard X-ray point sources around the
position of the cluster -- these are likely AGN. Cross-match them
with your optical galaxy catalog, and plot them on your cluster
color-magnitude diagram. What can you say about their cluster
membership? What kind of galaxies do they live in? Make sure you
look individually at every cross-matched source brighter than
Step IV: Write up your
results in the form of a short thesis proposal to your advisor.
Your write-up should have the following format (page lengths do
not include figures; they should go at the end of your writeup):
- Intro: why is this an interesting problem? what has been
done before? (2+ pages)
- Techniques: What you did in your data mining task.
Describe how the clusters were chosen (and why those criteria
were used). Describe your color-mag plots for each cluster,
and explain how you defined the red sequence and how you
defined spectroscopically confirmed cluster members. (2+
- Results: Did you find AGN in clusters? How many? With
what certainty? What kind of galaxies were they in? Other than
AGN in clusters, what kind of objects did you find? What were
the limitations of what you did? (2+ pages)
- Future steps: What would the next step be? What data
would you need to actually get a better handle on the problem?
What interesting scientific questions could you tackle? (2+
- Appendix: Your writeup should also include an
appendix which describes, for each cluster:
- The cluster redshift you determined.
- A figure showing the cluster CMD with X-ray sources and
spectrsocopically confirmed cluster galaxies overlaid.
- A listing of all X-ray sources brighter than r=20.5,
listing SDSS position, r mag, g-r color, SDSS-Chandra
crossmatch separation, and (if available) spectroscopic
- The writeup should be ~10 pages of single-spaced text,
not including figures, appendix, and references.
The archival data retrieval and analysis for steps I, II, and III
will be done in class. The writeup (step IV) should be your own
- As a start to your literature search, I'd
suggest looking at a couple of ApJ papers by Martini et al (2006,
This of course should be your starting point, not the end of
your reading! From there, you can look at papers they cite as
well as papers that cite them, and then continue on down
the literature trail.
- We will be working at distances were cosmological effects
will start becoming noticeable. So don't use Hubble's law to
get "a distance". Instead, use the cluster redshift to get
luminosity distances and angular size distances for
calculating spatial sizes and luminosities, respectively.
Here's how do do that in astropy:
import WMAP9 as cosmo
= cosmo.luminosity_distance(redshift) # in Mpc
= cosmo.angular_diameter_distance(redshift) # in Mpc
appmag - 5*np.log10(DL.value) + 5
r_arcsec * DA.value / 206265.