Target Selection for the DEEP2 Redshift Survey

1
Target Selection for the DEEP2 Redshift Survey
Atlanta, Jan. 2003

• Jeffrey Newman
• for the DEEP2 Team

2
The DEEP2 Collaboration

• Team Members
• U.C. Berkeley M. Davis (PI), A. Coil, M. Cooper,
  B. Gerke, R. Yan, C. Conroy
• U.C. Santa Cruz S. Faber (Co-PI), D. Koo, P.
  Guhathakurta, D. Phillips, C. Willmer, B. Weiner,
  R. Schiavon, K. Noeske, A. Metevier, L. Lin, N.
  Konidaris, G. Graves
• U. Hawaii N. Kaiser, G. Luppino
• LBNL J. Newman, D. Madgwick
• U. Pitt. A. Connolly JPL P. Eisenhardt
• Princeton D. Finkbeiner Keck G. Wirth
• K survey (Caltech) K. Bundy, C. Conselice, R.
  Ellis, P. Eisenhardt

3
Scientific Goals of the DEEP2 Redshift Survey

• 1) Characterize the properties of galaxies
  (colors, sizes, linewidths, luminosities, etc.)
  at z1 for comparison to z0 samples
• 2) Study the clustering statistics (2- and
  3-point correlations) of galaxies as a function
  of their properties, illuminating the nature of
  the galaxy bias
• 3) Measure the small-scale thermal motions of
  galaxies at z1, providing a measure of ?m and
  galaxy bias
• 4) Determine the apparent velocity functions of
  galaxies and clusters at high redshift, providing
  constraints on fundamental cosmological
  parameters such as wP/?DE

4
DEEP2 in brief

• 4 Fields 14 17 52 30 (includes Groth Survey
  Strip), 16 52 34 55, 23 30 00
  00, and 02 30 00 00
• Field dimensions 30 by 120 (15 ? 120 for
  Groth field) 20x80x1300 h-1 Mpc comoving
• Primary Redshift Range z0.7-1.4, preselected
  using BRI photometry to eliminate galaxies with
  zlt0.7 (otherwise would make up 60 of the sample)
• Spectroscopy 60,000 1-hour, R5000 DEIMOS
  slitmask spectra of targets to a limit of
  RAB24.1
• High resolution allows us to split the OII
  doublet and obtain kinematic information for a
  large fraction of targets

Keck access 80 nights of UC time over 3
years gt45 of the way done!!!
5
Comparison between DEEP2 and local surveys
SDSS
2dFGRS
LCRS
DEEP2
z0
CFASSRS
PSCZ
z1
6
DEEP2 vs. previous surveys of distant galaxies
Galaxies found in large numbers well beyond z 1
7
Testing our BRI color selection

• We now have thousands of redshifts in the
  Extended Groth strip, for which we attempt to
  observe targets regardless of their color. This
  allows us to test the impact of our target
  selection (which was tuned based on 300 zs) on
  the survey.

8
Our cuts are highly successful!

• By applying a relatively simple BRI color cut,
  we have a sample that is 13 zlt0.75, vs. gt60
  with no cut whereas only 3 of objects that were
  rejected are at zgt0.75 (5 gt0.7).

9
What to remember about the DEEP2 sample

• 1) In all fields but EGS, zlt0.7 galaxies are
  excluded by our preselection effectively. We
  will need to investigate what types of objects we
  miss in more detail (mostly just mag. errors?).
• 2) The sample is magnitude-limited in CFHT RAB.
  This corresponds to restframe 4000 Å at z0.7,
  2800 Å by z1.4 .
• 3) Even the brightest galaxies with red
  rest-frame U-B will go past the magnitude limit
  at zgt1.1-1.2 .
• 4) We observe only 70 of possible targets due
  to slitmask-making constraints (spectra cannot be
  allowed to overlap on the detector). Even with
  adaptive tiling, 3 slitlets, etc., we do worse
  than this in dense regions.

10
DEEP2 Color-magnitude diagrams vs redshift
0.65-0.8
Faint red galaxies are missing seen also in
local samples
0.8-1.0
1.0-1.2
lt 1.2
Weiner et al. 2003, Willmer et al. 2003
11
Update on Marc Davis...

• Marc suffered a stroke in late June his recovery
  and rehabilitation is ongoing, and will continue
  through much of this year.
• He is now visiting campus most afternoons,
  attending team meetings, reading email, etc. His
  participation increases every week, but the top
  priority for now remains rehab.

12
Redshift Distribution of Current Data
Color cut is working very well!
We are currently measuring redshifts for 80 of
targets in one hour of spectroscopy. Many DEEP2
failures are at zgt1.5 .
13
CFHT BRI photometry is quite effective for
selecting objects with zgt0.7

• Plotted at right are the trajectories galaxies
  observed at z0 would take in our color-color
  space as a function of redshift. Diamonds are
  plotted every 0.2 in z the transition from zlt0.7
  to zgt0.7 is marked by the change from dotted to
  solid lines.
• A simple curve (nearly parallel to the
  reddening vector) can be used to distinguish
  low-redshift from high-redshift objects. If we
  do not apply such a color cut, half the galaxies
  we observe would be at zlt0.7 (and our sample
  would be much more dilute as a result).

14
Advantages of a high-dispersion survey
Blue curve fraction of sky flux in a 100 Å
window coming from sky lines (as opposed to
continuum) Red curve fraction of pixels in that
same window that are on sky lines, for a 1200
l/mm grating. Most pixels have low background
effective OH suppression
15
Advantages of a high-dispersion survey
The high resolution used for DEEP2 observations
yields well-resolved linewidths for all objects,
and rotation curves as a free byproduct for
thousands. Shown are four 2d spectra exhibiting
resolved OII emission and the derived circular
velocity Vc(r).
16
Coordinated observations ofthe Extended Groth
Strip
MIPS, IRAC (Deep)
Background 2 x 2 deg from POSS
MIPS, IRAC (Med)
DEEP2/DEIMOS Spectra
DEEP2/CFHT B,R,I
WFPC2/Groth V,I
SCUBA
In this field, we will
XMM Chandra
- apply no zgt0.7 color cut
- survey half the area, but with
twice the mask density of other fields
17
DEEP2 has been made possible by DEIMOS, a new
instrument on Keck II
DEIMOS (PI Faber) and Keck provide a unique
combination of wide-field multiplexing (up to 160
slitlets over a 16x4 field), high resolution
(R5000), spectral range (2600 Å at highest
resolution), and collecting area.
18
We are assembling detailed portraits of the local
Universe but how did it reach its current state?
The DEEP2 Faint Galaxy Redshift Survey has been
designed to answer these questions by studying
both galaxies and large-scale structure at z1 in
detail for comparison to the present-day
Universe. In this way, we can see cosmic
evolution in action.
19
Redshift Surveys have been vital to our
understanding of the Universe

• The original CfA Survey (2,000 redshifts,
  25/clear night) revealed the filamentary nature
  of the galaxy distribution, motivating some of
  the first numerical simulations of LSS
• Surveys more than 100 times as large have been
  completed since (e.g. 2dFGRS) or are underway
  (SDSS).
• These new surveys provide strong constraints on
  statistics of the local (z lt 0.2) galaxy
  distribution, of great interest for testing
  cosmological models
• Large surveys also allow measurements of the
  distributions of galaxy properties (e.g.
  luminosity function, velocity function, stellar
  mass function), in addition to correlations
  amongst properties of individual objects

20
DEEP2 in Summary

• 3.5 sq. degrees in 4 fields surveyed
• ?60,000 targets, selected based on BRI photometry
  to have zgt0.7
• gt50,000 redshifts
• 6106 h-3 Mpc3
• One-hour exposures
• RAB ? 24.1 mag
• R5000 slitmask spectra linewidths for ? 70,
  spatially-resolved kinematics in gt20

Keck access 80 nights of UC time over 3 years
Observing season April-October
21
Surveys of distant galaxies can constrain both
galaxy formation and cosmology...

• The evolution of large-scale structure depends
  strongly on the underlying cosmology.
• By comparing the universe at high redshift to
  what is seen locally, many unique cosmological
  tests can be performed while simultaneously the
  evolution of galaxies may be studied!

22
What is left to be done after WMAP?

• Standard paradigm contains
• Dark matter ?m 0.27-0.07
• Dark energy 70
• Spectral index ns 0.99-0.04
• Equation of state parameter for DE wlt-0.8
• Unfinished business
• What is the meaning of such a mixture?
• Formation history of galaxies and clusters
• Better constraints, possible evolution of w

23
Survey strategy imaging

• We have obtained deep CFHT 12k imaging in three
  bands (BRI) to allow photometric pre-selection of
  targets with zgt0.7 otherwise, the majority of
  objects observed would be at lower z. The
  imaging is complete and fully reduced.

A 200? ? 200? BRI image from one of our fields
Photo-z preselection of targets
24
Slitmask spectroscopy
Using custom-milled slitmasks with DEIMOS we are
obtaining spectra of 120 targets at a time. A
total of 480 slitmasks will be required for the
survey we can tilt slits up to 30 degrees to
obtain rotation curves.
25
Masks tiled across a 42x28 CFHT pointing
26
Multipass target selection

• On a given mask, we cannot allow spectra from
  different objects to overlap - so tend to
  undersample dense regions (like clusters!)
• To ameliorate this, we overlap successive masks
  on the sky with an adaptive tiling, giving
  galaxies excluded by neighbors extra chances to
  be observed in secondary passes
• In the figure to right, the first-pass region of
  each mask is drawn, and the objects are
  color-coded by mask. Most objects on each mask
  are in its primary region, but a few may be found
  outside.
• 70 of all selected objects observed

27
DEIMOS slit masks and detector

• Slit masks are curved to match the focal plane
  and imaged onto an array of 8 2k?4k MIT-LL CCDs
• Readout time for full array (150 MB!) is 40
  seconds (16 amplifier mode)

28
First spectroscopy of DEEP2 masks

• Each slitmask has 120 objects over an 8k x8k
  array. The average slit length is 5 with a gap
  of 0.5 between slits. We tilt slits up to 30
  degrees to trace the long axis of a galaxy.

29
DEIMOS reduced data
Right A small percentage of one mask an OII
playground!
Left We will obtain thousands of well-resolved
rotation curves
Below Analysis of a tilted slitlet reduced data
above, raw data below. We routinely achieve
Poisson-limited sky subtraction in most cases.
30
A fully automated reduction pipeline
SDSS spectral pipeline code by Schlegel et al.
allowed us to rapidly develop a full 2d and 1d
spectral reduction pipeline that is completely
automated
A few percent of one DEEP2 mask, rectified,
flat-fielded, CR cleaned, wavelength-rectified,
and sky subtracted. Note the resolved OII
doublets. Shown is a small group of galaxies
with velocity dispersion ? ? 250 km/s at z?1.
Note the clean residuals of sky lines!
31
Status of the DEEP2 Survey

• DEIMOS commissioning began June 2002 under clear
  skies and was extremely successful.
• DEEP2 observing campaign began in July 2002. At
  the end of 3 semesters of the 6 planned, we have
  completed 48 of the survey slitmasks (plus 8
  masks for KTRS)!
• Observations complete mid 2005 (we hope)
• Analysis complete late 2006

32
Early results and current work include

• Spectroscopic classification of galaxies at z1
  (Madgwick et al., accepted)
• The dependence of clustering (?) on galaxy
  properties
• (Coil et al., submitted)
• Mock catalogs for DEEP2 (Yan et al., submitted
  Yan et al., in prep)
• Detection and membership determination for
  clusters and groups of galaxies (Gerke et al., in
  prep)
• Satellite galaxy dynamics (Conroy et al., in
  prep)
• Luminosity function evolution (Willmer et al., in
  prep)

33
Early results and current work include

• Resolved kinematics of galaxies (Cooper et al.,
  in prep)
• The dependence of galaxy properties on
  environment (Cooper/Gerke/Madgwick et al., in
  prep)
• Unresolved kinematics of galaxies linewidths
  (Weiner et al., in prep)
• Stellar populations in red galaxies (Schiavon et
  al., in prep)
• Angular correlations in the DEEP2 photometric
  sample (Coil et al., in prep)
• OII emission in red galaxies (Konidaris et al.,
  in prep)

And many more to come!!!
34
Principal Component Analysis (PCA)
PCA allows us to define a minimum set of
eigenspectra that span most of the variance in
our sample. The most influential component
primarily quantifies the strength of OII 3727.

Madgwick et al. 2003 astro-ph/0305587
35
PCA for classification
The strength of the first PCA eigenvalue alone
provides an effective means for determining
spectral types of galaxies, as seen in the
stacked spectra of galaxies split according to
this value.
36
Galaxy colors can also be used for
classification...
PCA allows us to classify galaxies based upon
their spectra however, we can also use our BRI
photometry, along with redshift, to derive
rest-frame broadband colors. Like at z0, the
distinction between early and late types is
readily apparent.
Weiner et al. 2003, Willmer et al. 2003
37
Clustering in DEEP2 First Redshift Maps
Projected maps of two DEEP2 pointings (of 13
total). Red early-type (from PCA).
38
2-point correlation function x(r)
x(r) measures the excess probability above random
of finding a galaxy in a volume dV at a distance
of r from a randomly chosen galaxy dPn dV
(1x(r) ) where n is the mean number density of
galaxies. x(r) measures the clustering in the
galaxy distribution. x(r) is known to follow a
power-law prescription locally x(r) (r0/r)g
with r05 Mpc/h and g1.8. r0 scale where the
probability of finding a galaxy pair is 2x
random In the DEEP2 survey we measure galaxy
clustering as a function of redshift, color,
spectral type and luminosity!
39
DEEP2 Faint Galaxy Redshift Survey

• Details, Scientific Goals of DEEP2
• Highlights of DEIMOS spectrograph
• Survey Status and Data Pipeline
• Science Topics in Progress

40
Dependence of galaxy properties on environment
The Voronoi volume of a galaxy is the amount of
space that is closer to that galaxy than any
other it provides a parameter-free measure of
the inverse number density of galaxies about any
object (cf. Marinoni et al. 2002). High z
resolution is required.
We can use this measure to study how galaxy
properties such as LF, color, spectral type, and
linewidth vary with environment in the DEEP2
sample (and compare with local surveys). For
instance, PCA emission-line galaxies are
preferentially found in low-density regions
Voronoi partition in 2 dimensions
Gerke et al., Cooper et al., in prep
41
Galaxy Groups and Clusters in DEEP2
Voronoi-based methods can also be used to
identify clusters and groups of galaxies
(Marinoni et al. 2002). We are currently
optimizing such techniques with mock catalogs,
and have begun producing DEEP2 group
catalogs. This will allow both the study of
group property distributions and of group vs.
field galaxies.
redabsorption-dominated
redpairs blueNgt2 size?log (?) ? log (halo
mass)
Gerke et al. 2004, in prep
42
Luminosity Function evolution
DEEP2 luminosity function measurements are well
underway. Good agreement with COMBO-17 in range
of overlap also LF as a function of color,
spectral type, etc.
Willmer et al. 2003
43
Luminosity-linewidth relations
Since we can measure both luminosities and
linewidths of DEEP2 galaxies, we can also explore
the relationship between the two and compare to
lower-z samples. Preliminary results suggest
the T-F relation becomes brighter at higher
redshift, in agreement with previous work (but
with much larger samples).
Weiner et al. 2004
44
Velocity dispersions of satellite galaxies
We can explore the potential wells of galaxies at
larger radii by examining the relative velocities
of faint neighbors of bright galaxies (ala Prada
et al. 2003). Preliminary tests on the data are
promising, and we are testing our ability to
reject interlopers with the mock catalogs of Yan
et al. (2003). We should have a sample of
hundreds of satellites by the end of DEEP2.
Conroy et al. 2004
45
DEEP2 and other Surveys

• DEEP2 fields are magnets for panchromatic study
  of galaxies, groups and clusters
• Comparison of groups found to Sunyaev-Zeldovich
  field survey maps
• Comparison to X-ray maps (Chandra proposal
  submitted)
• Comparison to weak-lensing mass maps
• Deep SIRTF imaging in 7 IR bands, as well as
  GALEX imaging in the UV.
• HST/ACS imaging, eventually
• Integrated picture of galaxies and clusters to
  z1.5 should allow us to test for the sorts of
  systematic effects that may already dwarf
  statistical uncertainties.

46
DEEP2 Conclusions

• DEIMOS observations began last July 5!
• DEEP2, in combination with local surveys (e.g.
  2dF, SDSS), will provide a variety of constraints
  
• Galaxy formation and evolution
• Galaxy clustering
• Measurements of cosmological parameters
• All spectra and results to be made public in
  timely fashion

47
Using DEEP2 for Cosmological TestsComoving
volume vs Redshift
Green different equation of state values wP/?
for ?m0.3 Volume varies by a factor of 3 at
Z1!! Not a small effect.
48
Update on Marc Davis...

• Marc suffered a stroke in late June his recovery
  and rehabilitation is ongoing, at his home.
• He is now visiting campus, attending team
  meetings, reading email, etc. His participation
  increases every week, but the top priority for
  now remains rehab.