The Formation of Massive Ellipticals

First, remember basic properties of elliptical galaxies, from ASTR 222:

Spheroidal, dominated by internal velocity dispersion, not much rotation.
Red colors, indicating old stellar populations with very little new star formation.
Form a tight relationship in color-magnitude diagram known as the red sequence.
Most commonly (but not completely) found in high density environments (updated Dressler 1980 plot from Houghton+ 2015)

What can we say about when, where, and how massive E's form?

Constraints from galaxy colors

Colors are red, which by themselves are only a weak constraint on age. But the low scatter in color along the red sequence provides an important constraint.

Imagine that a galaxy forms in a single burst of star formation, and then ages passively from that point on. Also imagine that galaxies form at a time t_form and that you observe them at a time t_obs. Finally, imagine that galaxies can form all at the same time, or randomly over time. Then the scatter in the red sequence depends on:

spread in formation age (synchonization)
rate at which galaxies redden with time
when you observe the galaxies

write this color scatter (Delta) as

where beta is a synchronization parameter that goes from 0 to 1:

beta=0: pure synchronization (all form exactly at t_form)
beta=1: random over range t_obs - t_form

Bower+ 92: Virgo and Coma clusters

observed scatter is ~ 0.04 magnitudes in color at z=0.

If no synchronization (beta=1), t_form>13 Gyr ago.

If synchronized (beta=0.1), t_form>7 Gyr ago.

Virgo and Coma red sequences from Bower+ 1992

The scatter at higher redshift gives better constraints, since you are looking back closer to the formation era, when colors are changing more rapidly.

Ellis+ 97: clusters at z=0.5

Three different galaxy clusters showed color scatter of ~ 0.07 magnitudes.
derived t_obs-t_form ~ 5 Gyr, unless formation is very highly synchronized (in three different clusters?)
in LCDM, z=0.5 is t_obs=8 Gyr, so t_form=3 Gyr, or z_form > 2.

Constraints from stellar populations

What about if we look at the stars in nearby galaxies and ask when they formed? Population synthesis studies of nearby elliptical galaxies by Thomas et al (2010) plot inferred star formation rate (y-axis) against time/redshift (x-axis) for elliptical galaxies of different masses (red: most massive, blue: less massive).

Two takeaways:

Massive ellipticals formed their stars very quickly and very early on in the universe's history
Downsizing: less massive ellipticals formed relatively later, and over a broader range in time.

Galaxy Formation versus Star Formation

Contradiction:

Colors and stellar pops say massive ellipticals formed their stars quickly at high redshift
High redshift galaxy samples say that massive ellipticals formed later:

How do we reconcile these two statements?

Remember: star formation and galaxy formation are different things!

From De Lucia and Blaizot (2007) -- stars may be old, but galaxies formed later:

Merger tree:

Assembly (blue) and star formation (green) history:

Ellipticals in the early universe

But massive ellipticals do exist at high redshift. Why are they hard to find?

Cimatti+ 2004: sample of ellipticals at z ~ 1.6-1.9, stellar masses ~ 10¹¹ Msun:

Spectra suggest ages of ~ 1-2 Gyr, so formation redshifts of z ~ 3-4:

These are estimated to be comparable in number to massive star forming galaxies at z~2 and could account for ~ 10% of the massive galaxies at z~0.

But most galaxy formation models find it hard to make this many massive galaxies this fast.

"Inside-out" Growth of Ellipticals

Massive ellipticals much more compact at high redshift. From Conselice 2014 ARAA:

So ellipticals seem to be growing "inside out" -- compact inner regions form first, they then accrete material at larger radius, growing in size. Two phase assembly:

Phase 1: high-redshift, rapid assembly, low angular momentum gas rich infall, starburst activity
Phase 2: ongoing accretion of less-bound, higher angular momentum material to grow in size. Must be "dry" (no gas) to preserve red spheroidal structure.

Lessons from field ellipticals at z=0

Field ellipticals often show morphological peculiarities and (sometimes) bluer colors and evidence for gas accretion:

NGC 474:

Centaurus A (nearest E):

Optical/HI (stars, neutral gas)	Spitzer (warm dust, tracing gas)

NGC 7252 (merger remnant -- E to be?)

Summary galaxy formation scenario:

Environment driven

high density, early collapse, galaxy formation at high redshift

lower density, later collapse, galaxy formation a continual process

Wet versus Dry merging

Gas-rich wet merging (high-z, or low-z field): starburst, possible regrowth of a disk in the spheroid
Gas-poor dry merging (clusters): galaxies grow while staying "red and dead"

Mass-driven ("downsizing")

things at the center of high density peaks form first, and also are (or become) the most massive objects (luminous Es)

lower mass things form later/continually

This is a schematic "starting point" -- contradictory in places, exceptions exist, and details are complicated and uncertain!