Faculty member: Sandra Faber
Participating students: Judy Cheng, Genevieve Graves

Note: This is just an example of several projects that Sandy is offering. See her for more information. The DEEP team (Faber, Koo, GuhaThakurta) is seeing 2-3 students for founding through the PhD.

1) Project name and description:

"Environments of red-sequence galaxies of different morphological
types in the Sloan Digital Sky Survey"

Description: A major new discovery of the past few
years for galaxies is COLOR BIMODALITY, in which galaxies
are grouped into two classes: blue, star-forming
galaxies and red-and-dead galaxies. The latter
are what we often classify as ellipticals and S0s. They can
be arranged in a magnitude (mass) sequence, and the question
arises, why has their star-formation stopped (or at least
greatly slowed down) when it continues in many other objects
of roughly the same mass?

A major problem with current work is that the red sequence
as identified just by red color is actually quite
inhomogeneous. To understand it better, we need to weed out
galaxies that are actually still star-forming
and only appear red because they contain dust. We also want
to discriminate pure E's from S0's, as the presence or absence
of a disk (as in S0's) is a powerful clue to whether a recent
merger has occurred (mergers destroy disks). Finally, we
want to identify galaxies that are basically red but may
still be undergoing nonzero (albeit low) star formation, as
these objects may be in transition and may give us clues on
how star formation ceases.

Theories for why SFR shuts down are basically of three types:
- One theory says that SFR stops because the supply of
gas falling onto galaxies slows down or stops. We
think that gas can only fall onto galaxies in the form of
cold clumps. Recall that visible galaxies are actually
embedded in much larger and more massive dark-matter
halos. Some simple physics says that gas in massive
halos above some threshold value must be hot (millions
of degrees). This could create a threshold HALO MASS
(around 10^12 Msol) above which gas cannot cool. Galaxies
would then shut down when their halos exceeded this mass.
This is called "massive-halo quenching."
- The second theory says that SFR shuts down due to the
formation of a black hole, which accretes gas and in
the process becomes an active galactic nucleus. Feedback
from the AGN heats the surrounding gas and drives it
out of the host galaxies, quenching SFR. What causes the
sudden formation of the black hole is not too clear; one
of the favorite theories is that black holes grow during
galaxy mergers, which simulations show can drive a lot
of gas to the center of a galaxy in a short time.
This is called "AGN quenching."
- The third theory says that SFR shuts down when a smaller
galaxy falls into a massive dark-matter halo above 10^12
Msol, which is filled with hot gas. Again, no more cold
gas can accrete onto the object. Such smaller galaxies
are called "satellites," and this is "satellite-quenching."

(After the initial quenching, there may also be a need to _keep_
gas hot to prevent it from falling in. AGN are often invoked
for this task, which is called "maintenance mode." Maintenance
mode is interesting but probably can't be studied using the data
base in this project, so we won't consider it further.)

Your project would build on the successful project last year by
Judy Cheng. Judy visually inspected 1400 SDSS red sequence
galaxies, which were provided to her by Jenny Graves. Jenny
weeded out all suspected star-forming galaxies and Seyfert galaxies
based on their emission line ratios. The sample was also picked to
have measured environments (i.e., are they in dense or sparse regions)
and to have GALEX UV images (as an additional clue to possible low-
level star formation). Hence, this first sample of galaxies that
Judy classified consisted of objects that we would conventionally
term E's and S0's, with a few Sa's and Sb's thrown in.

Judy visually classified all 1400 galaxies into three bulge/disk
categories: no disk (E's), strong disk (S0's, Sa's, Sb's),
and intermediate. She also noted any non-axisymmetric distortions
such as spiral arms, bars, dust, H II regions,and tidal interactions
due to mergers. This catalog will be valuable in itself, but she
then went on to compare her visual classifications with machine
parameters derived by Luc Simard (at Dominion Astrophysical Observatory)
using his automated galaxy profile-fitting program GIM2D. GIM2D fits
bulge and disk components to a galaxy image and also determines a
residual smoothness parameter (actually, UNsmoothness), which
represents deviations from the smooth bulge and disk fit. Judy found
quite good agreement between her visual estimates and Luc's automatic
parameters. This opens the way to using Luc's catalog of over 100,000
galaxies, now that we have an intuitive idea of what Luc's numbers mean.

Two environment measures are available: a) SDSS neighbor densities
computed by David Hogg, which key on the number of very nearby
neighbors, and b) environments computed by Michael Cooper,
member of the DEEP2 survey, using his delta_3 parameter. Delta_3
has been tuned to work well with DEEP2 survey data and keys
on the number of neighbors in a somewhat larger volume
than the Hogg environments.

Evidently, at the completion of the data phase of this project,
we will have assembled a number of the key parameters
that are needed to test which mechanism(s) play a role in
galaxy quenching---environments are a surrogate for dark-halo mass;
ongoing mergers can be seen in the SDSS images and
the presence or absence of a disk constrains whether a
merger could have occurred in the recent past; UV GALEX photometry
is a sensitive measure of residual low-level star-formation;
and SDSS spectra (and possibly images) can reveal AGN. We
can't say yet exactly how the analysis will play out, but it
seems to be a rich data set that is likely to yield important
conclusions.

Your project would launch from this jumping-off place. Three phases
are envisioned:
- Part 1: visually classify the star-forming galaxies and the
Seyfert galaxies that Jenny threw out of the sample on the
first pass, to gain a full understanding of all red sequence
morphological types. This part would involve classifying
several hundred objects, fewer than Judy did but still
substantial. Adding your sample to Judy's would yield roughly
2000 visually classified galaxies.
- Part 2: analyze the environments of these 2000
galaxies, sorted by galaxy type. What mass dark-matter
halos do the various morphologies inhabit? What fraction of
quenched red-sequence are in halos above and below the
critical halo mass? Where do mergers occur? Where do
AGNs occur?
- Part 3: if time permits, extend the analysis to the 100,000
galaxies measured by Simard. These have Cooper environments
if not Hogg values, and the statistics would be immensely
strengthened.

2) Scientific importance

The origin of red sequence galaxies is a central question in
galaxy evolution. Red sequence galaxies are one extreme
end of the Hubble sequence, so understanding their origin may
unlock clues to understanding the reasons for the whole Hubble
sequence. Related to this is the role that black holes and AGN
play in galaxy evolution: are they central or are they just
decoration? This project could shed light on this important
question.

3) Expected tasks for first year

- Learn how to download and inspect Sloan images.
- Prepare a database of fundamental galaxy properties
for your objects.
- Visually classify several hundred galaxies. Once you get
into it, this will take you only a few days.
- Compare your classifications to numerical morphological
parameters from GIM2D and the Sloan database.
- Devise a way (if possible) to predict your visual results
from machine-made numerical morphological parameters.

Summer tasks:

- Analyze the frequency of various morphologiess as a function of
magnitude, radius, environment, and other parameters. Can
we predict what kind of galaxies are found where?
- From previous step, try to isolate groups of galaxies that
are quenched by different mechanisms. Example: small galaxies
that possess disks and reside in massive DM halos are candidates
for satellite quenching.
- Possibly compare your results with models of galaxy formation,
which include quenching via the above mechanisms. Do these
models match your counts of different kinds of galaxies in
different environments?
- Write two papers. One would present your catalog of classifications,
the other would describe the statistical analysis of morphologies.

4) Skills needed or to be cultivated.

On this project you will learn:

- About forefront models for galaxy formation
- How to access the Sloan Digital Sky Survey database
- How to visually classify galaxy images with particular
attention to disks, ongoing star-formation, and signs
of ongoing or recent mergers and AGNs.
- Other quantitative tools used to characterize galaxy
morphologies, such as axis ratio, concentration, inclination.
- About related work such as stellar population modeling and
spectral analysis of both near and distant galaxies.
- The workings of the DEEP2 survey on distant galaxies.
- How to write a major paper.

5) Prospects for publication

100% if work is completed. Also co-authorship on other
related papers in which your work is used.

6) Whether the project will extend into the summer and into
second year

- Will certainly extend into summer. Is very open ended.
Will inevitably lead to further questions if student
and professor are both happy with the work being done.

7) Whether funding for summer and/or second year would be
available, assuming student makes good progress

- Milestones will be provided in Fall and Winter quarters.
If you complete successfully, summer employment after first
year is guaranteed. Further support beyond that depends
on success of DEEP2 survey grant proposals and your interests
and aptitudes.