1. Vandana Desai (Spitzer Science Center)
The Dirt on Dry Mergers
Vandana Desai (Spitzer Science Center)
Arjun Dey (NOAO), Emma Cohen (Caltech), Tom Soifer (SSC, Caltech), Emeric Le Floc’h (Saclay)

2. How and when did elliptical galaxies form?
Naive Expectation:
Most massive galaxies form last, and should be young.
Reality:
Most massive galaxies have old stellar populations.
Solution:
Form stars early, assemble them late in dry mergers.
time
The overarching motivation for this work is to understand how and when elliptical galaxies form. We all know that in hierarchical models of galaxy formation, the most massive halos grow through merging of less massive halos. Here’s a schematic diagram from Wechsler et al. to illustrate the formation of a galaxy-sized halo, with the high-z universe up here and the low-z universe down here.
The most naïve expectation from looking at this diagram is that the most massive galaxies should form last and should be young.
However, we know from lots and lots of observations that, in reality, local elliptical galaxies are red and dead, and have old stellar populations.
Dry mergers represent a solution to this apparent problem. A dry merger occurs between two galaxies of roughly equal mass, neither of which has any gas. Thus, when they merge, there is no gas to be driven towards the central regions to form new stars. If elliptical galaxies formed via dry mergers, then their stellar populations could be old, but their mass could be assembled more recently.
However, in hierarchical models of galaxy formation, a galaxy builds up its mass through mergers of smaller units. In this scenario, massive ellipticals should be among the last populations to form. While the cold dark matter scenario is very appealing, it has been a challenge to reconcile the expected recent formation times of ellipticals with their old stellar populations.
Dry mergers are one solution to this problem. Perhaps ellipticals form via the recent merger of two smaller red, dead gasless galaxies that formed their stars at high redshift. Then you could explain the old stellar populations while fitting in with the hierarchical merger scenario.
Wechsler et al. 2002

3. Do dry mergers really happen?
Surely we can find examples somewhere in 9 square degrees.
So the burning question is “Do dry mergers really happen?”. Here is where the connection to NOAO’s anniversary comes in. Our friends Buell and Arjun have constructed a wonderful survey using NOAO facilities.
The NOAO Deep Wide-Field Survey (NDWFS for short) is a whopping 9 square degree survey, ideal for looking for rare events, such as dry mergers between really massive local ellipticals. The primary ground-based optical imaging was carried out in a wide B filter, R, and I. There is also a substantial amount of K-and imaging. This is a great public resource at your disposal.
Bootes field

4. Red, luminous, early-type, nearby
van Dokkum et al. 2005
NDWFS+MUSYC
Red, luminous, early-type, nearby
Pieter van dokkum published a highly-cited paper in 2005 that looked for dry mergers in the Bootes field. He also threw in the much smaller MUSYC field, but most of his sample comes from Bootes.
I’m going to explain his sample a little bit, since our work builds uses the same one. His experiment was relatively straightforward. The initial color selection was based purely on optical colors. He identified galaxies that lie on the red sequence at around z = 0.1, and only galaxies greater than Lstar made the sample. Out of this sample of about 126 red galaxies, 86 had early-type morphologies. These can be divided into two groups: Those with significant tidal distortion, which are dry merger candidates, and those with no evidence of a recent tidal distortion. Those are just isolated red, dead ellipticals. We’ll call them the control sample.
61 Dry merger Candidates
25 Control Galaxies

5. Dry Merger Candidates van dokkum 2005
Here are some examples of what he found. These are three-band mages. The tidal features are red and broad, evidence that the merger did not involve much gas and was dry.
van dokkum 2005

6. Most of today’s field ellipticals formed via (major) dry merger.
van dokkum et al. 2005
71%
of early-type galaxies are morphologically disturbed
Most of today’s field ellipticals formed via (major) dry merger.
So the conclusion of this work is that 71% of early-type, red, luminous galaxies are morphologically disturbed, and that most of today’s field ellipticals formed via dry mergers.

7. Are they really dry? van dokkum 2005
So the question we had was simply, “Are these dry merger candidates really dry?” After all, they were selected in the optical. Maybe they are red in part because they are dusty?
van dokkum 2005

8. Look in the Infrared
So, luckily, NOAO facilities were not the only ones to look at the Bootes field. Most of the 9 square degrees was also observed by Spitzer with both IRAC and MIPS. So we made multiwavelength SEDs of each of the dry merger candidates. Here’s a close-up view of one such SED. You can see the two GALEX points here. The optical NOAO points here. The Spitzer IRAC here, the MIPS 24, and the MIPS 70. Most of these weren’t observed at 70 micron, but I’m just showing you an example that included the full data set.

9. Look in the Infrared
PAH?
Here is the same SED with the simplest of old stellar population models overplotted. This is just a 3500K blackbody fit to the first three IRAC bands. The purpose of this is to illustrate the huge excesses that we observe at 8, 24, and 70 microns. Note that the 8 micron IRAC band samples a strong PAH complex, which is often seen in starforming galaxies. Just looking at the SEDs, many within the sample show a jump at 8 microns like this.
Significant Infrared excess over stellar blackbody.

10. Spitzer-8 indicates heating by star formation!
Many groups have used some version of an IRAC color-color diagram to classify galaxy SEDs. This plot shows the color versus the 5.8 versus 8.0 color. Starforming galaxies should have a strong PAH at 8 micron, so they will appear to the right on this plot. Galaxies that have a strong AGN tend to fall in this wedge, as shown by several folks, including Stern et al using observations of the Bootes field. The filled symbols are the dry merger candidates. The empty diamonds are the control sample. The size of the stars is correlated to the observed 24 micron flux density, and the yellow stars have the largest mid-infrared excesses. The bottom line is that most are not powerful AGN.

11. Look in the Infrared
Chary & Elbaz 2001
Given that most objects avoid the AGN wedge, we simply fit the excess 24 micron flux with a starforming template. Here we show a template from Chary and Elbaz (2001). For this source, this template gives a star formation rate of almost 7 solar masses per year. That’s pretty much the largest one, which I showed you because it’s easiest to see the excesses.
Excess can be (mostly) accounted for by star formation.

12. The NDWFS Bootes Sample
< 1 Msun/yr
Here are a lot more examples, which are probably impossible to see. This shows the sample of dry merger candidates in the bootes field. The left shows the subset that were detected at 24 micron. The right shows the subset that were not detected at 24 micron. Given the detection limits of the Bootes Spitzer MIPS survey and the redshift of the sample, we can say that the objects in the right-hand mosaic have star formation rates less than 1 solar mass per year. The objects in the left-hand mosaic have a variety of star formation rates, ranging from roughly zero to about 7 solar masses per year.
Spitzer MIPS-24 detected
undetected

13. Spitzer-24 reveals residual star formation in ~25% of sample.
This is a good walking-around number. About 25% of them show star formation rates greater than 1 solar mass per year. While that’s not a totally insignificant amount of star formation for a galaxy like the Milky Way, these are the most massive, most luminous red galaxies in the local unvierse. van Dokkum found only 126 in 9 square degrees. They have stellar masses of around 10^11 solar masses. So this amount of star formation really represents residual star formation.

14. Dare we infer gas masses?
After extrapolating, etc:
Gass mass~108 Msun
Moist major merger or accretion of dwarf galaxy?
OK, so our original question was whether these things were dry. We’ll, we’ve seen dust emission, and inferred a star formation rate. We can go even further and estimate a gas mass. This is very uncertain, because it requires extrapolating the SED to very long wavelenghts based on the best-fitting template fit at moderate wavelengths to calculate a dust mass. And then converting that dust mass into a gas mass. But we soldiered on. These galaxies have on order 10^8 solar masses in gas.
So the question is, is this consistent with the dry merger scenario. 10^8 solar masses does not constitute a large fraction of the stellar mass of the galaxy. So, we could be looking a a moist major merger instead of a dry merger. Or, are we looking at a recent accretion of a dwarf galaxy? Unfortunately, distinguishing between these two scenarios with the present data set is pretty difficult.

15. This star formation frosting is likely triggered by mergers.
No tidal features
Weak features
Number
Strong features
Here I show histograms of the derived SFRs as a function of the strength of the tidal features. The solid lines represent sources with MIPS detections, while the dotted lines represent sources without MIPS detections, and are just limits. You can see that the sources without tidal features have no sources with large star formation rates. All of the sources with SFRs > 1 solar mass per year have some morphological evidence for merger activity. This suggests that the merger may be responsible for triggering the star formation that we see in ellipticals.
Ongoing mergers
SFR (Msun/yr)

16. Summary
One quarter of dry merger candidates have enough gas to fuel SFR > 1 Msun/yr.
This corresponds to residual star formation on top of a massive old stellar population.
Dry mergers or minor mergers?
This residual star formation is linked to signatures of merger activity.

17. Spitzer-8 indicates heating by star formation!