1. 1 A Multiband Imager for Magellan Christopher Stubbs Department of Physics Departme nt of Astronomy Harvard University cstubbs@fas.harvard.edu Christopher Stubbs Department of Physics Departme nt of Astronomy Harvard University cstubbs@fas.harvard.edu

2. 2 Conceptual Design Pass  > 7000 A Pass  > 5500 A Pass  > 8500 A i z r g Common shutter

3. 3 Existing “standard” passbands U B V R I

4. 4 Pan-Starrs bands g r i z y 4000 A break at z of 0.2 0.5 0.8 1.0 1.25 1.5

5. 5CollaboratorsCollaborators Christopher Stubbs, CfA & Harvard PhysicsChristopher Stubbs, CfA & Harvard Physics Melissa Franklin, Harvard Physics DeptMelissa Franklin, Harvard Physics Dept Tony Stark, CfATony Stark, CfA John Geary, CfAJohn Geary, CfA South Pole Telescope/Dark Energy Survey collaboration: J. Carlstrom, J. Mohr...South Pole Telescope/Dark Energy Survey collaboration: J. Carlstrom, J. Mohr... Alan Uomoto, OCIW, has been very supportive Christopher Stubbs, CfA & Harvard PhysicsChristopher Stubbs, CfA & Harvard Physics Melissa Franklin, Harvard Physics DeptMelissa Franklin, Harvard Physics Dept Tony Stark, CfATony Stark, CfA John Geary, CfAJohn Geary, CfA South Pole Telescope/Dark Energy Survey collaboration: J. Carlstrom, J. Mohr...South Pole Telescope/Dark Energy Survey collaboration: J. Carlstrom, J. Mohr... Alan Uomoto, OCIW, has been very supportive

6. 6 Science Opportunities Supernova followup observationsSupernova followup observations Type Ia and type II Sne as cosmological probes Requires multiband images, multiple epochs Photometric redshifts of clustersPhotometric redshifts of clusters 4 band imaging over modest field Transient followupTransient followup Evolution of SED for GRBs Microlensing light curves Planetary occultationsPlanetary occultations Multiband data useful for discrimination Supernova followup observationsSupernova followup observations Type Ia and type II Sne as cosmological probes Requires multiband images, multiple epochs Photometric redshifts of clustersPhotometric redshifts of clusters 4 band imaging over modest field Transient followupTransient followup Evolution of SED for GRBs Microlensing light curves Planetary occultationsPlanetary occultations Multiband data useful for discrimination

7. 7 Cosmology from SZ Clusters Multiple projects now funded and under way to use SZ effect to detect galaxy clusters, down to ~2 x 10 14 M solarMultiple projects now funded and under way to use SZ effect to detect galaxy clusters, down to ~2 x 10 14 M solar Expectation is ~ 7 clusters per sq degreeExpectation is ~ 7 clusters per sq degree South Pole Telescope will map 4000 sq deg; 29,000 clusters SZ signal strength z-independentSZ signal strength z-independent Optical observations needed for zOptical observations needed for z Spectroscopy – SALT Wide field multiband survey – DEC 2009 Targeted moderate field multiband observations Multiple projects now funded and under way to use SZ effect to detect galaxy clusters, down to ~2 x 10 14 M solarMultiple projects now funded and under way to use SZ effect to detect galaxy clusters, down to ~2 x 10 14 M solar Expectation is ~ 7 clusters per sq degreeExpectation is ~ 7 clusters per sq degree South Pole Telescope will map 4000 sq deg; 29,000 clusters SZ signal strength z-independentSZ signal strength z-independent Optical observations needed for zOptical observations needed for z Spectroscopy – SALT Wide field multiband survey – DEC 2009 Targeted moderate field multiband observations

8. 8 South Pole Telescope (SPT) Survey Bolometric focal plane, 1000 elements, and 10m aperture telescope.Bolometric focal plane, 1000 elements, and 10m aperture telescope. Will map Southern extragalactic skyWill map Southern extragalactic sky RA from 20 hrs to 7 hrs DEC from –30 to –75 degrees Currently funded by NSF polar programs, under constructionCurrently funded by NSF polar programs, under construction Early 2007 first light.Early 2007 first light. Bolometric focal plane, 1000 elements, and 10m aperture telescope.Bolometric focal plane, 1000 elements, and 10m aperture telescope. Will map Southern extragalactic skyWill map Southern extragalactic sky RA from 20 hrs to 7 hrs DEC from –30 to –75 degrees Currently funded by NSF polar programs, under constructionCurrently funded by NSF polar programs, under construction Early 2007 first light.Early 2007 first light.

9. 9 Anticipated Cluster z-distribution 90% within z<1.2 50% within z<0.5 Carlstrom, Holder and Reese Ann Rev Astron Astrophys, 2002

10. 10 Photometric Redshift for Clusters Photo-z’s for individual galaxies tend to have scatter of  z /(1+z)~0.03, but with a few “catastrophic” outliers.Photo-z’s for individual galaxies tend to have scatter of  z /(1+z)~0.03, but with a few “catastrophic” outliers. Combination of morphology, magnitude, color and location can be used to establish cluster’s redshift.Combination of morphology, magnitude, color and location can be used to establish cluster’s redshift. Robust statistics can be used to eliminate “outliers”.Robust statistics can be used to eliminate “outliers”. Photo-z’s for individual galaxies tend to have scatter of  z /(1+z)~0.03, but with a few “catastrophic” outliers.Photo-z’s for individual galaxies tend to have scatter of  z /(1+z)~0.03, but with a few “catastrophic” outliers. Combination of morphology, magnitude, color and location can be used to establish cluster’s redshift.Combination of morphology, magnitude, color and location can be used to establish cluster’s redshift. Robust statistics can be used to eliminate “outliers”.Robust statistics can be used to eliminate “outliers”.

11. 11 Photometric Redshifts in SDSS bands Blanton et al, astro-ph/0205243

12. 12 Early-type galaxies in SDSS bands g-r 0.7 r-i 0.3 r-z 0.8 g-r 0.7 r-i 0.3 r-z 0.8 Bernardi et al, astro-ph/031629 g 20.7 r 20.0 i 19.7 z 19.2

13. 13 Brodwin et al, astro-ph/0310038 Relevant Magnitudes are 18<r<24

14. 14 m AB griz Median redshift (r mag) (r mag) 233586320870 0.52 to 23 rd 23.57821080022000.73 24180500190055000.85 24.545012004900130001.01 Time (sec) to reach SNR=10 Extended source, ABmag in 2.2 arcsec aperture Dark time, 0.8 arcsec, airmass=1.2, scaled to Magellan and high-  http://rpm.cfht.hawaii.edu/~megacam/diet/DIET.rpm Galaxy colors roughly follow contours of constant integration Shaded boxes are DEC target 10  magnitudes. We should get ½ the clusters (those with z<0.5) in 60 sec

15. 15 Clusters per unit telescope time Z range N Time (seconds) 0 – 0.2 15 15*60 = 900 0.2 – 0.4 30 30*60 = 1800 0.4 – 0.6 30 30*60 = 1800 0.6 – 0.8 15 15*200 = 3000 0.8 – 1.0 10 10*600 = 6000 1.0 – 1.2 9 1.2 – 1.4 4 15*1000= 15000 1.4 – 1.6 2 totals115 ~4 hrs for 100, z<1 4.2 hours for 15 more, z>1

16. 16 Return on Magellan Time Investment Could get 200 clusters out to z~1, or 100 clusters out to z of 1.5, in 1 night. Fifteen dark nights on Magellan could produce photo-z’s for 3000 clusters out to z=1, about 10% of the total expected from SPT. Could get 200 clusters out to z~1, or 100 clusters out to z of 1.5, in 1 night. Fifteen dark nights on Magellan could produce photo-z’s for 3000 clusters out to z=1, about 10% of the total expected from SPT.

17. 17 Conceptual Optical Design Exists Doublet field flattener (2 aspheric surfaces) Field stop Identical triplets (all spherical) T. Stark, CfA

18. 18 Optical Performance Plate scale is 0.062 arcsec per 15  m pixel FOV is 4.1 x 4.1 arcminutes (~ 700 kpc at z=0.3) 80% encircled energy in ~0.15 arcsec: Plate scale is 0.062 arcsec per 15  m pixel FOV is 4.1 x 4.1 arcminutes (~ 700 kpc at z=0.3) 80% encircled energy in ~0.15 arcsec:

19. 19 Tightly coupled software/observing Take Image 1 30 sec Analyze Image: flatten, WCS, sextractor Galactic reddening corr. Produce z,  z OK? Offset Take Image 2 30 sec Slew to next target Offset if appropriate More images

20. 20 Rapid Readout and High Efficiency Integrated SW reduces wasted time Binning 2x2 gives 0.124 arcsec/pix, Single amplifier output per 2K x 4K chip 8 sec readout @ 250Kpix/sec Effective telescope time multiplier: For “balanced” exposures, sequential images take T series = 4(t exp +t readout )x N frames Parallel imager takes (with 0.8 throughput degradation) T parallel = (t exp /0.8 + t readout ) x N frames For t exp >>t readout total time is reduced by a factor of ~3 Integrated SW reduces wasted time Binning 2x2 gives 0.124 arcsec/pix, Single amplifier output per 2K x 4K chip 8 sec readout @ 250Kpix/sec Effective telescope time multiplier: For “balanced” exposures, sequential images take T series = 4(t exp +t readout )x N frames Parallel imager takes (with 0.8 throughput degradation) T parallel = (t exp /0.8 + t readout ) x N frames For t exp >>t readout total time is reduced by a factor of ~3

21. 21 Controlling Systematics Our design is readily baffledOur design is readily baffled Can use both field stop and pupil stop Suppresses stray and scattered light Better flatfielding Single common shutter near pupilSingle common shutter near pupil Reduced shutter artifacts Flux ratios with a single pointing and 2-3 exposures, under all conditions!Flux ratios with a single pointing and 2-3 exposures, under all conditions! Even with patchy cloud cover, get Poisson-limited colors. Our design is readily baffledOur design is readily baffled Can use both field stop and pupil stop Suppresses stray and scattered light Better flatfielding Single common shutter near pupilSingle common shutter near pupil Reduced shutter artifacts Flux ratios with a single pointing and 2-3 exposures, under all conditions!Flux ratios with a single pointing and 2-3 exposures, under all conditions! Even with patchy cloud cover, get Poisson-limited colors.

22. 22 Complementarity with Dark Energy Camera Multiband Camera: 16 arcmin 2, 6.5m aperture: - Get 100 clusters to z~1 in 4 hours. - Get 100 clusters to z~1 in 4 hours. - Can be on the sky at start of SPT survey - Can be on the sky at start of SPT survey Dark Energy Camera: 3 sq deg, single band, 4m aperture: - Survey approach on CTIO 4m gets 3 sq deg x 7 clusters/sq deg = 21 clusters per hour, or 84 clusters in 4 hours. - Survey approach on CTIO 4m gets 3 sq deg x 7 clusters/sq deg = 21 clusters per hour, or 84 clusters in 4 hours. - Delivers other science as well: weak lensing, SN detection... - Delivers other science as well: weak lensing, SN detection... Multiband imager’s cluster hunting advantage is 100/84 = 1.2 Multiband imager’s cluster hunting advantage is 100/84 = 1.2 Does not take into account seeing advantage on Magellan Does not take into account cloud-immunity of multiband imager If SPT survey falls short of flux goals, multiband advantage increases Multiband camera better suited to chasing z>1 clusters. Multiband Camera: 16 arcmin 2, 6.5m aperture: - Get 100 clusters to z~1 in 4 hours. - Get 100 clusters to z~1 in 4 hours. - Can be on the sky at start of SPT survey - Can be on the sky at start of SPT survey Dark Energy Camera: 3 sq deg, single band, 4m aperture: - Survey approach on CTIO 4m gets 3 sq deg x 7 clusters/sq deg = 21 clusters per hour, or 84 clusters in 4 hours. - Survey approach on CTIO 4m gets 3 sq deg x 7 clusters/sq deg = 21 clusters per hour, or 84 clusters in 4 hours. - Delivers other science as well: weak lensing, SN detection... - Delivers other science as well: weak lensing, SN detection... Multiband imager’s cluster hunting advantage is 100/84 = 1.2 Multiband imager’s cluster hunting advantage is 100/84 = 1.2 Does not take into account seeing advantage on Magellan Does not take into account cloud-immunity of multiband imager If SPT survey falls short of flux goals, multiband advantage increases Multiband camera better suited to chasing z>1 clusters.

23. 23 Two-stage plan 2007 – 2009:2007 – 2009: SPT in operation SPT in operation Use multiband camera to image ~2000 clusters Use multiband camera to image ~2000 clusters Initial cluster count vs. z results Initial cluster count vs. z results Use camera for SN followup as well... Use camera for SN followup as well... Post-2009:Post-2009: Dark Energy Camera on Blanco 4m (if $) Dark Energy Camera on Blanco 4m (if $) Deep wide DEC survey will find SNe Deep wide DEC survey will find SNe Follow SN light curves with multiband camera Follow SN light curves with multiband camera Chase higher-z clusters with multiband camera Chase higher-z clusters with multiband camera 2007 – 2009:2007 – 2009: SPT in operation SPT in operation Use multiband camera to image ~2000 clusters Use multiband camera to image ~2000 clusters Initial cluster count vs. z results Initial cluster count vs. z results Use camera for SN followup as well... Use camera for SN followup as well... Post-2009:Post-2009: Dark Energy Camera on Blanco 4m (if $) Dark Energy Camera on Blanco 4m (if $) Deep wide DEC survey will find SNe Deep wide DEC survey will find SNe Follow SN light curves with multiband camera Follow SN light curves with multiband camera Chase higher-z clusters with multiband camera Chase higher-z clusters with multiband camera

24. 24StatusStatus Rough conceptual design doneRough conceptual design done Proof-of-concept optical design doneProof-of-concept optical design done Detectors are in handDetectors are in hand Lincoln labs 2K x 4K, 15  m pixels Some epitaxial, some high-resistivity Readout electronics require replicationReadout electronics require replication MegaCam architecture and board set Machine shop capacity availableMachine shop capacity available Rough conceptual design doneRough conceptual design done Proof-of-concept optical design doneProof-of-concept optical design done Detectors are in handDetectors are in hand Lincoln labs 2K x 4K, 15  m pixels Some epitaxial, some high-resistivity Readout electronics require replicationReadout electronics require replication MegaCam architecture and board set Machine shop capacity availableMachine shop capacity available

25. 25 Task list ModelingModeling Slew vs. expose tradeoffs vs. redshift and richness Photo-z determination HardwareHardware Finish design work, order optics Mechanical design, fabrication Electronics and detector optimization System integration, testing... SoftwareSoftware Scripts that connect Sextractor to photoz codes Test with SDSS and IMACS data Integrate with instrument ProposalsProposals Plan is to submit NSF ATI proposal Nov 1. ModelingModeling Slew vs. expose tradeoffs vs. redshift and richness Photo-z determination HardwareHardware Finish design work, order optics Mechanical design, fabrication Electronics and detector optimization System integration, testing... SoftwareSoftware Scripts that connect Sextractor to photoz codes Test with SDSS and IMACS data Integrate with instrument ProposalsProposals Plan is to submit NSF ATI proposal Nov 1.

26. 26SummarySummary A multiband imager for Magellan makes sense:A multiband imager for Magellan makes sense: Multiple science drivers Unique capability Quality photometry Conceptual design completedConceptual design completed Detectors on hand, optics straightforwardDetectors on hand, optics straightforward Design needs to be refined, and optimal observing strategy devisedDesign needs to be refined, and optimal observing strategy devised Software is a key aspect.Software is a key aspect. Your comments & participation welcome!Your comments & participation welcome! A multiband imager for Magellan makes sense:A multiband imager for Magellan makes sense: Multiple science drivers Unique capability Quality photometry Conceptual design completedConceptual design completed Detectors on hand, optics straightforwardDetectors on hand, optics straightforward Design needs to be refined, and optimal observing strategy devisedDesign needs to be refined, and optimal observing strategy devised Software is a key aspect.Software is a key aspect. Your comments & participation welcome!Your comments & participation welcome!

27. 27 Backup Slides and residual clutter

28. 28 Cluster luminosity function DM= 38.2 implies L* ~ 17.1 AB r mag at z = 0.1  (M) Christlein and Zabludoff, astro-ph/0304031

29. 29 Solid: w=  1 Dotted: w=  0.6 Shortdash: w=  0.2 Longdash: open CDM

30. 30 Limiting Magnitude Limiting Magnitude ½ L* cluster galaxies at redshift 4000A break leaving blue filter½ L* cluster galaxies at redshift 4000A break leaving blue filter  g,r,i,z = 22.8,23.4,24.0,23.3  Complete cluster catalog Galaxy catalog completenessGalaxy catalog completeness  g,r,i,z = 22.8,23.4,24.0,23.6  Simple selection function Blue galaxy photo-z at faint magsBlue galaxy photo-z at faint mags  g,r,i,z = 24.0,24.0,24.0,23.6  Photo-z for angular power spectra and weak lensing ½ L* cluster galaxies at redshift 4000A break leaving blue filter½ L* cluster galaxies at redshift 4000A break leaving blue filter  g,r,i,z = 22.8,23.4,24.0,23.3  Complete cluster catalog Galaxy catalog completenessGalaxy catalog completeness  g,r,i,z = 22.8,23.4,24.0,23.6  Simple selection function Blue galaxy photo-z at faint magsBlue galaxy photo-z at faint mags  g,r,i,z = 24.0,24.0,24.0,23.6  Photo-z for angular power spectra and weak lensing AB Mag of ½ L*