1. The Evolving Luminosity Function of Red Galaxies
Michael J.I. Brown
Buell Jannuzi, Arjun Dey, Kate Brand, Mark Brodwin, Peter Eisenhardt, Andrew Benson, Darren Croton
Massive Galaxies Over Cosmic Time 2
The formation, assembly and evolution of galaxies are key unresolved issues in modern astrophysics, and have been identified as such by both the Australian and Amercian decadal reivew. Our framework for understanding galaxy evolution is provided by simulation and models, but beyond low redshift these remain largely untested by robust observations.
Today I will discuss my recent observational studies of galaxy assembly, which have measured the assembly of galaxies with unparelleled precision.

2. Red Galaxies Ideal for tests of simulations & models.
Contain the bulk of the stars at low redshift.
Include the most massive low redshift galaxies.
Little growth from recent star formation.
Image stolen from SDSS

3. Bootes Surveys NDWFS I-band IRAC 3.6m NOAO Deep Wide-field Survey
Spitzer IRAC Shallow Survey
AGES
Spitzer MIPS Bootes
XBootes
FLAMEX
GALEX
NDWFS I-band
IRAC 3.6m

4. Photometric Redshifts
Larger volumes than spec-z surveys.
Photo-zs accurate to +/-0.05 at I<22.
Luminosities &
colours from SED model fits

5. Red Galaxy Selection
Selection criterion falls between CMR and blue cloud.
Similar selection used in the recent literature.
Colour
As mentioned previously, red galaxies tend to lie along a well defined locus.
Most red galaxies lie along a well defined locus, so our measurements are not hugely depending on our selection cut
We choose a selection cut that allows direct comparison with the literature.
Magnitude

6. 39599 red galaxies Brown et al. (2006), ApJ, accepted astro-ph/0609584
41333 galaxies, this compares to 1000 galaxies in DEEP2 and 5000 galaxies in COMBO-17
Unlike DEEP2 we account for the colour evolution of red galaxy old stellar populations.
Unlike COMBO-17, we have verified the accuracy of our redshifts with spectra of ~10% of our entire sample.
Unlike many surveys we account for the faint, yet significant flux which fall outside of our aperture.
We should have smaller random uncertainties and smaller systematic errors than the prior literature.
Brown et al. (2006),
ApJ, accepted
astro-ph/

7. The Luminosity Function
Red galaxies were brighter in
the past.
Galaxies
per unit
volume
To quantify if blue galaxies are being turned into red galaxies, we measure the optical light being emitted by the ensemble, the total population, of red galaxies.
Brown et al. (2006),
ApJ, accepted
astro-ph/

8. Luminosity Density Stellar pop. model overestimates jB at z~1
Stellar mass in red
population increases.
Truncation of SF in blue galaxies.
Unlike previous studies we see a slight increase in the B-band luminosity density
Still we are below the simple model by a factor of 2 at z=1,
It would appear that galaxies are entering the red seq at z<1.
The question is how. Is it via mergers of blue galaxies fading onto the red seq
Brown et al. (2006),
ApJ, accepted
astro-ph/

9. The Luminosity Function
Measure the absolute magnitude were the space density of red galaxies is 10^-3.5 per cubic.

10. ~4L* Red Galaxies Steady evolution of luminosity with z
Differs slightly from stellar pop. model
Growth via mergers,
but not rapid growth.
Unlike other measures of red galaxy evolution, this one shows good agreement between various surveys at z<0.8. This reason being that it is only measuring the brightest redddest galaxies - so the selection effects and incompleteness have modest effects on the results.
Brown et al. (2006),
ApJ, accepted
astro-ph/

11. Summary Red galaxy assembly over the past 7 billion years:
Stellar mass within the red galaxy population doubles.
Star forming blue galaxies are being transformed.
The most luminous galaxies were assembled at z>1.
Some simulations overestimate z<1 galaxy assembly.
Key Remaining questions.
What truncates star formation in galaxies?
When were massive red galaxies assembled?
What are the progenitors of today’s most massive galaxies?
In contrast to many of the models

12. The z=0.5 Luminosity Function
Measure the absolute magnitude were the space density of red galaxies is 10^-3.5 per cubic.

13. The z=0.9 Luminosity Function
Measure the absolute magnitude were the space density of red galaxies is 10^-3.5 per cubic.

14. Luminosities
To measure the fluxes of red galaxies, we use 4” aperture photometry.
This image shows z=0.8, and the 4” aperture photometry appears to capture a large portion of the total flux.
The size-luminosity relation of red galaxies is a predictable function of luminosity (and redshift)

15. Luminosity-Size
This plot shows the difference between 4” and 6” aperture photometry as a function of lum.
The dots are our red galaxies while the lines a models derived from the size-lum.
While a 4” diameter aperture is large for high z galaxies, it clearly doesn’t capture all the flux.

16. Colour Luminosity (SDSS, D. Hogg)
Moderate luminosity red galaxies could be assembled via mergers with other galaxies. The could also be produced if star formation in a blue galaxy is truncated. The luminosity of stars goes roughly as their mass to the power of 4, so their lifetimes are on the order of mass to the power of 3. The very luminous blue stars in star forming galaxies therefore have very short lifetimes, and they go supernova, or become whitedwarfs, on timescales less than 1 Gyr.
No bright blue galaxies. Brightest red galaxies can’t be formed at this redshift by fading blue galaxies.
Luminosity
(SDSS, D. Hogg)

17. Observed Galaxy Mergers
While this is evidence for ongoing assembly. It is difficult to determine a precise assembly history from such galaxy merger candidates.
Van Dokkum (2005)
Timescales? Significance?