Spitzer Cosmic Far-IR Background Project

PI: Bruce Grossan(1)

Co-I: George F. Smoot(2)

In the graphic above, the microwave sky at bottom illustrates the matter in the very early universe, the second image shows the seeds of structure in dark matter (simulation), third image shows the far-Infra-red sky, a background light of galaxies being born in bursts of star formation, and the top image shows a fully-formed galaxy as might be seen today. (Copyright BCCP/B.Grossan/R.Nocera 2008; artwork by Rosemary Nocera)

Galaxies Do Not Form at Random- They Form in Structures. Can We Measure That Structure, and What Might It Tell Us About How Galaxies Are Formed?


Far-IR map of Lockman Hole	Power Spectrum of Galaxy Structure

Most recent astro-ph paper: http://arxiv.org/abs/astro-ph/0604512

PHASE I or Spitzer Cycle 1 Project
PHASE II or Spitzer Cycle 3 New Observations Project

What are we doing?

Learning about a different kind of cosmic background - the far-IR background (~160 microns).
Learning about the structure of the universe by looking (mostly) at light from galaxies at z ~ 1 during their burst of star formation
Applying our expertise in mapmaking to Far-IR maps

I. OUR MAP:
Here is the first Spitzer large-area map, of the Lockman hole:
Spitzer Lockman 160 micron map

II. MAKING OUR MAP BETTER:
When this map was planned, little time or consideration was given to mapping the background emission. Lots of streaks (stripes) in the vertical direction are evident.

Now, with a little knowledge of background mapmaking, we went and got just a tiny bit of extra data:
cycle3data

... and this allowed us to make a rather better map by simple processing of the timeline data:

elflockdr1map

Looks better.

We can also apply some fourier domain filtering of the image data (as opposed to the timeline data as above) to get a still better map:

fourier-filtered elflock center

Now notice that in this map (cropped map rotated relative to the map above - sorry) there is no structure along the lines of declination (vertical in the previous map, slanting right in this map). Big improvement.

III. The Science:
There are at least two extraordinary things about this picture; it was taken by the most sensitive instrument ever in this band and it is extremely large - it covers more than 14 contiguous square degrees. This image or map could only have been made with the Multiband Imaging Photometer for Spitzer, the MIPS instrument aboard the Spitzer Space Telescope . This instrument has an unprecedented ability to make maps of huge swaths of they sky at IR wavelengths up to 160 micron wavelength - in the far IR. The sky at these wavelengths glows with what is believed to be the aggregate emission from dusty galaxies out to red shift 1.2 or so; at this wavelength even with the powerful Spitzer MIPS instrument the sky is mostly this unresolved diffuse or background emission.

Predictions before 2003 (Perrotta et al. 2003) suggested that increased power in larger scale structures will be present in analysis of the structure of the Far-IR background. This would be the signature of galaxy clustering - never before seen in the far IR. In layperson's terms, this means that this map is not smooth; the patterns of bright and faint are arranged in clumps of certain sizes, which we call more generally structure. The largest clumps or structures are bright and obvious, but small structure is faint and not obvious. The significance of this structure is that it shows us how matter was distributed throughout the universe when star formation took place. This tells us how matter was distributed at the time these star-forming galaxies were formed. We know a great deal about how matter was distributed in the early universe from the Cosmic microwave background; matter is distributed very differently when galaxies were formed.

In Spitzer cycle 1 we examined the largest available maps with low foreground contamination, the Lockman Hole SWIRE observations, and analyzed the resulting 6 square degree map, the white rectangle in the map above.

THE STRUCTURE OF THE UNIVERSE
(as told in star-forming galaxies)

The power spectrum of the indicated sub-section looks like this:
Lockman Hole Power Spectrum

The solid dots give the raw power spectrum. The open circles give the power spectrum corrected for instrument response. The noise in the spectrum is the solid line at bottom. The lowest frequency points are the signal from the IR Cirrus foreground, due to dust in our galaxy. This can be subtracted off to show only the contribution from the CFIB...

The CFIB Spectrum, that is, with the foreground cirrus subtracted off and corrected for instrumental effects:

s11cfibspec+lagache

Previous spectra of the CFIB had found only data consistent with a flat spectrum, without any structure. This is really exciting because of the following theoretical power spectrum published before 2003 - Perrotta et al. (2003):

clustering prediction

The above figure shows a simulated power spectrum of galaxies layed down at random (Dole) plus foreground cirrus. YOu see the steep cirrus hits the flat-ish spectrum at mid-k, then decends with instrument effects.

WE CLEARLY SEE THE CLUSTERING SIGNATURE - A LOW-K EXCESS - IN OUR CFIB POWER SPECTRUM.

Project Phase II or Spitzer Cycle 3 Original Observations:

It is extremely well known that background observations are best done with maximized cross-linking, that is, with scan paths that cross each other at large angles. Look again at the Lockman Hole 160 micron Map, above. There appears to be some vertical structure in the background, or if you look at the scan map,

The Spitzer Lockman hole map was produced with this scan pattern:
Spitzer Lockman Scan Pattern

you can see this vertical strucutre is obviously related to the scan pattern. Note that, except for the small region at top midle (verification scans) , ALL of the scan lines are parallel!!!!!! Although in our paper we have done our best to correct and simulate the effects of this structure in our data, we can do much better with additional observations - these were awarded to us during Spitzer Cycle 3 and should be performed in Jan 2007. There will be 8 cross-liinking AORS or 8 cross-linking scans with 4 approximate repeats of each of them. There will also be a single total power mode observation to nail down the absolute flux level of the map and which we will use to inter-calibrate some more of the scans.

Institutional Affiliations:
(1) Eureka Scientific/LBNL, (2) LBNL/UC Berkeley Physics/BCCP

Bruce Grossan
email: try first name + underscore + last name at lbl dot gov to communicate via electronic post. [anti-spam format address]
510.486.5489