20 Nov. 2006 Jean-Paul KNEIB - prospective spatial PNG 1 A wide field imager for dark energy … and more ! SNAP-L Jean-Paul KNEIB LAM, Marseille, France

20 Nov. 2006 Jean-Paul KNEIB - prospective spatial PNG

1

A wide field imager for dark energy

… and more !

SNAP-L

Jean-Paul KNEIBLAM, Marseille, France


2

The SNAP-L Mission

SNAP-L is a ~2m telescope with a wide field optical/near-IR camera and a 3D optical/near-IR spectrograph.

SNAP-L is a project led by the Department of Energy (US particle physicist community) started in ~1999

with international partners: France: through the spectrograph development and

scientific expertise (SN, WL) [INSU+IN2P3] Sweden: SN science Canada: WL science Other countries interested … and possibilities to have a

stronger contribution in the project.


3

SNAP-L: the concept

A 2 meter class telescope, 3 mirror anastigmatic designProvide a wide field flat focal plane, FOV > 0.7 square degreesCovering ~350 to ~1700 nmOn a L2 orbit for stability and low background


4

Imager visible + IR pixel scale = 0.10 arcsec/pixel (visible),

0.17 arcsec/pixel (NIR)36 4kX4k CCDs [0.35-1]m 0.5 Gigapixels,36 2kX2k HgCdTe [1-1.7] m

spectrographe 3D IFU slicer visible+IRR=100-200

3’’x3’’ , [0.35-1.7]m

9 filters6 visibles3 IR

SNAP-L: focal plane


5

R&D on optical/IR Detectors

0

10

20

30

40

50

60

70

80

90

100

600 800 1000 1200 1400 1600 1800

Wavelength (nm)

QE (%)Rockwell #103

Raytheon #141

SNAP req.

0

10

20

30

40

50

60

70

80

90

100

200 400 600 800 1000 1200

Wavelength (nm)

QE (%)

CCD 86135-7-7HyVis H2RG-32-ALN-03i

SNAP req.

1

10

100

1 10 100Number of Fowler pairs

Read noise (e-)

Rockwell detector #103, 300 sec, 140 K

Raytheon detector #09A, 300 sec, 100 K

SNAP req.

Important R&D funded by DOE for SNAPOn the detectors have reached the SNAP requirements.

Latest IR detector should go on the new WFPC3 camera to be installed on HST during SM4.


6

SNAP IFU slicer spectrometer

IR path

Entrance point

Pupil & slit

mirrorSlicer

Prism

DetectorCamera

Collimator

Visible IR

Wavelength coverage (m) 0.35-0.98 0.98-1.70

Field of view 3.0" 6.0" 3.0" 6.0"

Spectral resolution, 70-200 70-100

Spatial resolution element (arc sec)

0.15 0.15

detectors LBL CCD10 m

HgCdTe18 m

Efficiency with OTA and QE

>50% >40%

• IFU concept based on slicer

• Compact and light (20x30x10 cm)

• Spectroscopy of SN and host in the same time

• Photometric calibration

• Spectro-z for photo-z calibration

• Demonstrator being developed at LAM


7

The SNAP-L Mission

SNAP-L is a dedicated mission to measure Dark Energy withSuperNovae and WL measurements (and possibly Baryon

Acoustic Oscillation).

SNAP-L will have a deep survey and a wide surveyBoth dedicated for SN and WL observation strategy but both

useful for « other sciences »

Key advantages of SNAP-L: PSF, image quality, stable photometry Wide field, Depth,Large wavelength coverage (both in visible and NIR with 9 filters),on board spectrograph.


8

Why going in space?

0.1

1

10

100

1000

10000

100000

0 0.5 1 1.5 2 2.5

Wavelength (m)

(Background

γ -1)sec

Zodiacal light

Paranal sky

D = 2 m = /D

D = 8 m = 0.3”

•0.1” angular resolution over wide field (0.7 sq.degree)•Near-infrared unfettered by atmospheric emission/absorption•Continuous, year-round observation of selected fields•Stability!


9

Space-based imaging vs ground

Space-based imaging has a significantly higher surface densityof resolved sources, which can probe the matter density powerspectrum at higher redshifts than will ever be feasible from theground.

GEMS COMBO-17 (Brown et al. 2003)

~ 100 galaxies per sq arcmin ~ 35 galaxies per sq arcmin


10

Survey Area(sq.deg) Depth(AB mag) ngal(arcmin-2) Ngal

Deep/SNe 10 30.4 250 107

Wide /WL 1000=>4000 28.1 100 108.5

Panoramic 7000-10000 26.7 40-50 109

SNAP SurveysSNAP SurveysP

oin

t S

ou

rce

- 3


11

SNAP Deep Survey

Base SNAP survey: 7.5 square degrees near North ecliptic pole

~3000x as large as ACS UDF to mAB=30.4 in nine optical and IR bands.

Provides ~150 epochs over 22 months (each to mAB=27.8) for time domain studies in all nine bands [SNe, AGNs]

GOODS Survey area

Hubble Deep Field


12

SNe Systematic Control

SNe observation strategy:

Goal: Observe 2000 hig-redshift SN in photometry and spectroscopy up to z~1.7How: 22-month survey covering 7.5 sq.degree, with 2400s exposure per field every 4 days.The 9 band photometry will allow to select SN candidates for spectroscopy, and ensure quality rest-frame photometry. 40% of the time is reserved for on-board spectroscopy, with a large fraction for z>1 SNe. SN redshift determine through the SiII broad line.

NICMOS on HST has shown that spectro-photometry calibration can achieve better than 1% error


13

SNAP Wide Area Survey

~1000 sq.deg. ‘wide’ survey the deep field, but discussion for extension to 4000 sq.deg.

Roughly 1 year for 1000°2 of observing time

Four dithered 500 second exposures at each location; sensitive to mAB=28.1 (point source)

Every field observed in all nine optical NIR filters

GOODS Survey area

Hubble Deep Field


14

WL 2-points stat: What is measured?WL 2-points stat: What is measured?

Mass power spectrum normalisationMass power spectrum normalisation

Slope of the power spectrumSlope of the power spectrum

Mean density parameterMean density parameter

Redshift of the sourcesRedshift of the sources

Ultimately Dark Energy constraintsUltimately Dark Energy constraints

<g<g22>~0.01 >~0.01 882 2 1.6 1.6 zzss

1.4 1.4 qq-(-(nn+2)/2+2)/2


15

Lensing Mass Map

-green countours: X-ray-Color blobs: optical/phot-z detection

3D Mapping of the mass distribution.

COSMOS field as an example.


16

Ground/Space comparison Shear Calibration

error estimate for a constant PSF : Ground 0.7’’ Space 0.1’’

m: is calibrated with ‘realistic’ image simulation m~5e-3.

m depends on PSF stability and ellipticity

QuickTime™ and aTIFF (LZW) decompressor

are needed to see this picture.

Waerbeke et al


17

Ground/Space Comparison

A space 4k sq.deg survey, is equivalent to a ground 20k sq. deg survey for similar photo-z bias.

Space photo-z bias should ~5 times better, a factor of 3 improvement in the FOM




18

Photometric Redshift

NIR Filters are crucial for photo-z accuracy and to reduce catastrophic failures (see Ilbert et al 2006)

Filter optimisation for photo-z in progress, possibility to include “U-band” filter.




19

The standard method -Results

CFHT-LS deep field photo-z show that SED templates needs to be optimized !!!

Ilbert et al 2006


20

Calibration - template optimization

Optimized templates

CFHT-LS optimize 4 templates with 2800 spectroscopic z

Initial templates

Need of spectro-zCalibration.

On-board spectrograph can measure redshift in parallel of the SN and WL survey (~50 000 spectro-z per year of WL observation AB<24.5)


21

Calibration - improvement

Calibration method is successful

to remove

systematics.

More spectro-z the better, feasibility is on progress but is looking good.


22

Dark Energy Constraints

Produce Good photo zUse 3 WL MethodsVery powerful


23

what SNAP can also

Dark energy: SNII Galaxy clustering /

baryon oscillations Galaxy clusters and their

clustering Strong lensing

Correlation with other surveys ISW, SZ, dark baryons

Non-dark energy science Galaxy evolution Quasars and AGN Solar system objects Nearby galaxies,

structure, stellar pops, globular clusters

High-z objects MW structure + stars


24

Strong Lensing with SNAP-L

Current example:SL2S: automatic search through the CFHT-LS for arcs and partial rings around elliptical galaxies (~40 candidates out of the first 28 sq.degree) + Follow-up with an ACS snapshot program.COSMOS: 1.5 strong lenses in 1.7 sq deg.

=> ~10-40 thousands strong lensing system in SNAP-L WL survey.

Cabanac et al 2006


25



Marshall et al 2006


26

ACS/grism, Keck/LRIS & VLT/FORS2 observations confirm z=5.83

UDF Can see Galaxies at z~6And has candidates up to z~8 - similarly SNAP-L will image these distant galaxies …


27

High-z galaxies

Stiavelli et al 2004

Expect ~100 000 z>7 galaxies in the SNAP-L SN surveys down to AB~29.

Unique way to map large scale structure at z>7 (faster than JWST) and find rare objects (QSOs, strong lenses, …)

SNAP-L can be seen as a survey telescope for JWST.


28

Probing the end of dark ages

z~3 quasars: 200 – 400 per sq. deg

Hundreds of z~6 quasars

Maybe 10 luminous quasars at z = 9 – 10?

Xiaohui Fan


29

Conclusion

SNAP is a well advanced concept (R&D well advanced and ready for integration) currently proposed in the NASA JDEM context, but JDEM contract is being re-discussed for an early launch (goal 2014).

SNAP is dedicated to dark energy and will provide at least 2 surveys (AB=30,28 point sources) for SN and WL but these can address many other sciences.

France (CNES) through the spectrograph contribution is well involved, and other participation might be possible to become a stronger partner (telescope, WL data center and analysis …)

Documents

20 Nov. 2006 Jean-Paul KNEIB - prospective spatial PNG 1 A wide field imager for dark energy … and more ! SNAP-L Jean-Paul KNEIB LAM, Marseille, France