SURVEY CHALLENGES
description
Transcript of SURVEY CHALLENGES
SURVEY CHALLENGES
Kirpal NandraImperial College London
With help from… Antonis Georgakakis, Elise Laird, James Aird, and the AEGIS team….
Chandra Surveys workshop : Survey Challenges
THE AEGIS SURVEY
• Chandra: 1.6 Ms over 0.5 deg2
• DEEP II spectroscopy• HST, Spitzer, VLA, GALEX,
CFHT LS blah blah blah“AEGIS” ApJL special issue (~20 papers accepted)
aegis.ucolick.org
Chandra Surveys workshop : Survey Challenges
WHAT HAVE WE LEARNED FROM X-RAY SURVEYS (1)?
• 0.5-5 keV XRB mostly resolved into AGN• Fainter soft sources high z starbursts• X-rays detect more AGN than optical• Additional Compton thick AGN provide 30
keV background• Evolution:
• Rapid evolution to z=1 as (1+z)3 like SFH• Low LX decline above z=1• All decline above z~3-4
Chandra Surveys workshop : Survey Challenges
WHAT HAVE WE LEARNED (2)?
• Type I/II fraction increases with LX
• But still many type II QSOs • The type I/II fraction decreases with z• AGN associated/coeval with star formation• Host galaxies: red, massive, bulge
dominated• Clustering like hosts, perhaps even more
Chandra Surveys workshop : Survey Challenges
SO WHAT ARE THE CHALLENGES?
• Cosmic variance i.e. large scale structure• Optical completeness and Identification• X-ray completeness and analysis issues• Statistical biases and small number stats• Selection methods and biases
Chandra Surveys workshop : Survey Challenges
COSMIC VARIANCE
Gilli et al. and Barger et al. noted z spikes in CDFs
Georgakakis et al. GWS ~300 arcmin2 Laird et al. AEGIS ~2000 arcmin2
Chandra Surveys workshop : Survey Challenges
Cosmic variance is a big problem, so we need wider
surveys, right?
WRONG…
Chandra Surveys workshop : Survey Challenges
Nandra et al. 2006
THE AGN COLOR-MAGNITUDE RELATION
Chandra Surveys workshop : Survey Challenges
Bright sources only (~Bootes limit)
DEEP VS WIDE
Nandra et al. sample (200ks)
Chandra Surveys workshop : Survey Challenges
DEEP VS WIDE
DEEP2 redshifts ~4 deg2
(Davis et al 2003)
Miyaji correlation functions
“Cosmic variance” is not a sin!
Chandra Surveys workshop : Survey Challenges
X-RAY SOURCE ENVIRONMENTS
vs. host luminosity vs. host colour
Comparing with galaxies samples same range of LSS
AEGIS: Georgakakis et al. (2006) also Coil talk
Chandra Surveys workshop : Survey Challenges
HIGH Z LUMINOSITY FUNCTION
z=3 luminosity function LX=1043-44.5 space density
Aird et al. (in prep + poster)Hasinger et al (2005)
Barger et al. (2005)
Completeness corrections are crucial
Chandra Surveys workshop : Survey Challenges
Chandra 2-8 keV L* (2-10) = 1.5E44 at z=1Evolves as (1+z)3 from z=0 to z=1
(Barger et al. 2005)
ASIDE ABOUT LOW Z EVOLUTION
Implies evolution more like (1+z)1-2
New hard X-ray data at z=0RXTE XSS 3-20 keV L* (2-10) = 5.E43 (Sazonov)Integral 20-40 keV L*(2-10)=6.0E43 (Beckmann)Integral 17-60 keV L*(2-10)=3.5E43 (Sazonov) Swift BAT 10-100 keV L*(2-10)=4.3E43
Chandra Surveys workshop : Survey Challenges
PHOTOMETRIC REDSHIFTS
Field: dz/(1+z) = 0.05 Failure rate = 5%X-ray: dz/(1+z) = 0.05 Catastrophic rate =10%
AEGIS/CFHTLS photo-z’s (Ilbert et al. 2006)
MUST ACCOUNT FOR PHOTO-Z ERRORS FAILURES IN ANALYSIS!
Chandra Surveys workshop : Survey Challenges
OPTICAL IDENTIFICATION
• Chance projections in AEIGS to I=25:7% IDs at 1.5”; 20% at 3”; 30% at 5”REAL IDs are optically fainter high z?
ID of SCUBA source GN11 (w/Alex Pope + Douglas Scott UBC)
HST/ACS IRAC 3.6mm radioMIPS 24mm
Alexander et al X-ray c/part Pope et al. c/part
MAJOR IMPACT ON NUMBER OF HIGH Z AGN/REIONIZATON
SCUBA AGN fraction may be lower than Alexander et al. (2005)
Chandra Surveys workshop : Survey Challenges
X-RAY INCOMPLETENESS
• X-ray images are• Highly inhomogeneous• In poisson regime
• Source detection “black box” (e.g. wavdetect)
• Detection inconsistent with sensitivity
• Eddington bias, poisson noise, incompleteness
• Embodied in sensitivity curve
Georgakakis et al., in prep
Chandra Surveys workshop : Survey Challenges
PERILS OF HARDNESS RATIOS
• HRs overestimate the absorbed fraction and NH because NH cannot be <0
• Especially bad at high z
• Simulations with =1.9 and dispersion 0.2 and NH=0!
• LBG at z=3 with HR=-1 actually has NH=1.5E23
Chandra Surveys workshop : Survey Challenges
A type II AGN is one without broad lines in its optical spectrum
A NOTE ON TYPE II QSOs
By this definition, no true type II QSOs have been found in X-ray surveys, as e.g. H is unobserved
(and if it is, it’s broad)
Chandra Surveys workshop : Survey Challenges
DO X-RAY SURVEYS FIND ALL AGN?
• Heckman et al. (2005) say OIII better at selecting local AGN than X-ray
• Steidel et al. (2002) found 70% of X-ray AGN at z=3 LBGs from spectroscopy
• Also one AGN X-ray undetected in 1 Ms• Sarajedini et al. (2006): 70% of optically variable nuclei
X-ray undetected (200ks Chandra)• AEGIS (Renbin Yan, Berkeley):
• 60% of X-ray sources have AGN line ratios• 10% have no OIII • Only 30% of line-ratio selected (candidate) AGN are X-ray sources!
Not to mention Spitzer selection… need multi- approachBut remember flux limits…
Chandra Surveys workshop : Survey Challenges
OTHER ISSUES
• Separating AGN and starbursts • Is it reasonable to assume Compton thick evolve like
unobscured• Is alpha_ox dependent on UV luminosity really?• How does variability affect SEDs. Dispersion?• Effects of variability effects on photoz?
Chandra Surveys workshop : Survey Challenges
WHAT HAVE WE LEARNED FROM X-RAY SURVEYS?
• 0.5-5 keV XRB mostly resolved into AGN• Fainter soft sources high z starbursts• X-rays detect more AGN than optical• Additional Compton thick AGN provide 30 keV background
• Rapid evolution to z=1 as (1+z)3 like SFH,, Low LX decline above z=1, All decline above z~3-4, little reionization contribution
• Type I/II fraction increases with LX
• But still many type II QSOs • The type I/II fraction decreases with z• AGN associated/coeval with star formation• Host galaxies: red, massive, bulge dominated• Clustering like hosts, perhaps even more
Chandra Surveys workshop : Survey Challenges
WHAT HAVE WE LEARNED FROM X-RAY SURVEYS?
• 0.5-5 keV XRB mostly resolved into AGN
Chandra Surveys workshop : Survey Challenges
WHAT ARE THE IMPORTANT QUESTIONS?
• How are AGN triggered?• Do they affect bulge/star formation, or vice versa? • What are the astrophysical processes implied by
obscuration?• Do X-rays tell us anything useful about star formation? • AGN contribution to the total luminosity of the universe
(c.f. stars)?• How do AGN affect the early universe?• Can clusters be used to constrain cosmology• What is the history of the baryons in the universe?
Chandra Surveys workshop : Survey Challenges
MSSTs
• Impact of environment on galaxies• Are AGN created by mergers?• History of accretion• Physics geometry and evolution of absorption• Physics and evolution of groups ad clusers• Effect of AGN feedback on galaxy hosts, groups
and clusters• Cluster mass function at high z and growth of
structure
Chandra Surveys workshop : Survey Challenges
ANEWs
• Generating and sharing data• All sky hard X-ray surveys• Followup of non X-ray obscured AGN• Deeper Chandra surveys• Better photoz• Large area cluster survey• Relevant Spitzer observations• Multiwavelength completeness corrections• Extend ultradeep surveys in areas with best NIR• X-ray surveys before Spitzer dies or JWST comes• Development of multivariate luminsity functions• Followup of variablity-selected objects
Chandra Surveys workshop : Survey Challenges
WAYS FORWARD?
• Better analysis of existing data is possible (and needed)
• Need to learn from each other• Archival and ground-based followup needed• And… more X-ray observations might be
justified… so what observations do we need?• Infinite depth, 4pi area?• In reality each problem defines its own
requirement in area-depth parameter space.
Chandra Surveys workshop : Survey Challenges
A (PROVOCATIVE) PROPOSAL
• Certain fields deep Spitzer and wide HST• Cosmos, AEGIS, GOODS, E-CDFS/GEMS (Chandra)
• There are only a few Spitzer wide fields • Bootes, FLS, SWIRE (XMM)
• Need Spitzer for complete AGN selection. Spitzer’s cryogen is running out
• Concentrate on making these fields better, and make all the data public
• Also need very wide fields:• Archival (2XMM, Champ, AXIS)• XMM slew survey of SDSS equatorial strip (250 deg2)? • All-sky (RXTE, BAT, Integral, eRosita)
Chandra Surveys workshop : Survey Challenges
CURRENT AND TARGET DEPTHS
Survey Area Depth Target Expo
CDF-N 0.1 2Ms 5Ms 3Ms
CDF-S 0.1 1Ms 5Ms 4Ms
E-CDFS 0.3 250ks 500ks 1.5Ms
AEGIS 0.6 200ks 500ks 2.4Ms
COSMOS 2(0.7) 200ks 200ks 5Ms
FLS 5 0 20ks 1.5Ms
Bootes 9 5ks 20ks 1.8Ms
SWIRE 50 Patchy 20ks 6Ms
SDSS Eq 250 0 2ks 6Ms