Bernard Fort (IAP)
CoI: Olivier Le Fevre & Jean-Paul Kneib (Marseille)Mireille Dantel-Fort (Obs. de Paris)
Christophe Alard, Yannick Mellier & Jean-François Sygnet (IAP)
Raphael Gavazzi & Genevieve Soucail (Toulouse)Remi Cabanac (CFHT)
The CFHTLS Strong Lensing survey
The CFHTLS: imaging / Vimos spectroscopic survey
• Which lenses can be seen?
• How many lenses?
• Which scientific objectives can be pursued?
How to find the lenses?
• A spectroscopic search of source/lens pairs
• Toward an automatic search of lens candidates
Overview of the presentation
The CFHTLS field D1
Petit film de J-C Cuillandre 1 minuteCf fichier sur la clé
• CFHTLS Deep Fields
4 x 1°, U, G, R, I, Z, IAB < 28.4 + VLT VIMOS Deep Spectroscopy 40 000 spectra, R = 5000
all objects with IAB< 24, <zspectro> = 0.76
. CFHTLS Wide Field
170 °, IAB < 24.5
CFHT Legacy Surveys/ VIMOS
2DFGRS/SDSS stops herez=0.5
z=0.6
z=0.7
z=0.8
z=0.9
z=0.6
z=1
z=1.3
z=1.1
z=1.2
160Mpc
30M
pc
VVDS 0226-04 cone with 6217 galaxies
(Olivier Lefevre et al 2004)
CFHTLS scientific drivers in Cosmology
• Formation and evolution of galaxies
• SN survey
• Weak Lensing (Mellier et coll. 2000) - recover the shape and amplitude of the power spectrum
- obtain cosmological parameters with a better accuracy (Jain, Berstein & Dolney astro-ph 0502243)
- test gravity on large scales (Bernardeau 2005 )
• Strong Lensing (Fort et coll. 2005) cosmological tests, by comparing
observations with simulations (many ref in Alcaniz et al 2005, astro-ph/0501026).
• number, ellipticity, profile, mass evolution of individual halos (cf CfA-Arizona-ST-Lens-Survey Falco et al 1999).
• Singly highly magnified sources ( > 3)
Magnification bias expl: Keeton 2005 Almaini et al 2005 astro-ph/0501169: mm sources bias)
distant galaxy or QSO
Observer
DM halos
nh = 0.025 halo/Mpc3
Zs > 2
the line of sight intersect
Nh> 400 halos.
A ground for optimism
Fraction of multiply lensed QSOs ~2.10-3
JVAS + Cosmic Lens All-Sky Survey 12/5000 distant radiosourcesare lensed by a foreground (E) galaxy(astro_ph/0211069, Browne et al; Chae 2002, 2004 , Chen et al, 2004, ApJ 607, L71)
<zlens> ~ 0.4, <Dsource> ~ 4 Gpc
no = 0.5 10-2 Mpc-3, < REinstein> ~ 1 arcsec
for SIS:
~ no < (REinstein)2 > Ds ~ 10-3
Various kind of strong lensing "events"
• galaxy (QSO)-galaxy strong lensing events
• multiple arc(let) systems in groups and clusters
• Singly highly magnified lens events (>3-5)
• Possible unexpected lens events (multi-plan events, dark lenses, cosmic
strings,..)
A HSF smple of galaxy- galaxy lenses: big Elliptical represents 2/3 of the lenses
Arcs around galaxies
20 lensed Lyman α background
galaxies for 2000 massive, E / bulge-dominated galaxies
with z>0.4,R<20,B-R>2.2
(APM survey: Willis et al, 2000)
10 – 30 % of all very distant sources are magnified with >10 (Keeton’s prediction 04 astro-ph/0405143)
CFHTLS highly magnified drop-out galaxies
>3
NWF
SIS
Omont et al. 2005
A CFHTLS gallery of giant arcs from Mellier 2005 discovered in the W3 field.
Giant arcs in clusters of galaxies
Schecter distribution L(z)/L*(z)of SIS halos
Faber-Jackson (Tully-Fisher) law
Observational relation (SLOAN) DM = f(*)
d/dz = n(E, z) (1+z)3 XS c.dt/dz
+
+
<XSNFW (z)>
Comoving number n NFW (z)(Mo and White 2000)
+
Estimate of the optical depth
CFHTLS: expected number of lens events (completude IAB<26.5)
• SL type predicted number observed (ref.)
• giant arcs ~ 0.1 / sq. degree ~ 1 / sq. degree (Mellier 2005) • arc(let) systems (groups, cluster) ~ 10-20 / sq. degree ~ ? • SGGL (mostly E) ~ 50-70 / sq. degree ~ many (HST, GOODS,..) • Singly H-magnified > 10-100 / sq. degree (?) ~ ? • others lenses ~ 0 - ? / sq. degree
10 times more giant arcs in CFHTLS than first CDM predictions (Bartelmann et al. 98) .
X-section depends on DM profile and geometry ellipticity, tri-axiality, source sizes, external shear,
merging (varying caustic length), central stellar mass,..(cf. Bartelmann et al 95, 02, et al., Oguri 02, Guo-Liang Li et al. 03,
Dalal et al. 04, Wambsgans et al. 03, Menegghetti et al. 04)
Simulations are the only way to calculate
properly for any peculiar statistical study
Giant arcs
- Frequency of lenses
- Distribution function:
image separation
redshift lenses
sources luminosity
...
.L
O
S
{ } ~ (Dol Dls/Dos) . grad
Dol
DosDls
geometry
DM distribution
Cosmological tests with arc statistics
Prob[zlense] with various parameters
zsource
m ,
mass evolution
varying
(Ofek, Rix & Maoz 2003)
a Weak Lensing Survey of 300 sq° and
a deep SNAP suvey (15 sq°)will give respectively
15000 and 4900 lense events
(Marshall, Blandford et al. 2004)
SNAP or Deep Universe Probe (DUNE) perspective
a HDF source
arc(let)s
Simulations of arcs around E-lenses
a HDF Elliptical seeing = 0.8 arcsec
circular source CFHTLS
arc(let)s
circular lens
HST
?
?(spiral like!)
OII 3127ÅZs=1.32
4000 ÅZL=0.63
A spectroscopic search of SG-G L
I
G- I
11.000 spectra are already available and candidates are there ( O. Le Fevre, 2005) !
The CFHTL spectroscopic survey (40 000 spectra)should give more candidates than the total number of known multiple QSOs!
First results with VIMOS
Searching arcs around galaxies
(GOODS: r = 0.1 arcsec, 0.3 arcsec < RE < 3 arcsec)
is equivalent to search arcs around groups
(CFHTLS: r = 0.8 arcsec, 3 arcsec < RE < 60 arcsec
same ratio RE / r
An automatic search of arc systems
CFHTLS: schematic software procedure
Eliminate stars & select multiplets(n, quad, triplet,..)
with same color (U-G,..)
Select arclet sytemsfrom generic properties
of lens systems
Test models and flag candidates
Sextractor catalogsU,G,R,I,Z
arclets geometryx, y, , dij, ..
seeing
Mag., deflector
radial pair
parallel pair
saddle pair
tripletquadruplet
circular pair
A selection of generic arc(let) systems
F02BiG 1530,13184
Multiple images candidates
It would have been difficult to recognize such arclet systemswhen the Einstein radius get smaller (10 -> 3 arcsec)
Spectroscopic confirmations are allways needed
Constraints
• Define a clear selection criteria for each lensing event (giant arcs: length, magnification, limiting surface brightness, zs,..).
• Most parameters depend on observational factors!
• Evaluate observational biases (completude, false detections..)
Working plan (2005-2007)
- define and test best algorithms for the automatic
search procedure on CFHTLS
- Test the robustness of the automatic detection:
number of possible missing events, false alarms
- Prepare the statistical lensing analysis
Immediate scientific targets
. Detect and model lens systems from groups
. Study new lens configuration (multi-plan lenses,dark lenses, ..)
. Find new arclets sytems in cluster lenses . Study the singly highly magnified
population of drop-out galaxies
Arcs
HST/ACS : Cosmos survey : Gavazzi, Van Waerbeke, Mellier, Fort
Conclusions
- It seems quite possible to develop an automatic search of lensing events with five colors per field.
- the methodology is tractable but the definition of clear selection criteria and the evaluation of systematic bias will be a real challenge.