Big BOSS The GroundBased Stage IV BAO Experiment

Big. BOSS: The Ground-Based Stage IV BAO Experiment Submitted to Astro 2010 Decadal Survey April, 2009

Big. BOSS: The Stage IV BAO Experiment The Questions Physics beyond the standard model Why 72% dark energy now? (It was less in the past, more in the future) Is the Universe geometrically flat? Is gravity the same on all scales? Imprint of non-gaussian fluctuations from early-Universe inflation!? These fluctuations of 1 part in 105. . . these ~unity fluctuations today gravitationally grow into. . . Universe at 300, 000 years old (CMB) Universe today (galaxy map)

BAO and dark energy What are baryon acoustic oscillations (BAO)? This sound wave can be used as a “standard ruler” Dark energy changes this apparent ruler size These fluctuations of 1 part in 105 gravitationally grow into. . . Universe at 300, 000 years old (CMB) . . . these ~unity fluctuations today Universe today (galaxy map)

Large Surveys: What’s the Future? Sensitivity to new physics scales as volume surveys -- # of modes Our observable Universe Volume mapped by SDSS + SDSS-II M. Blanton for SDSS Volume to be mapped by SDSS-III (ca. 2015) JDEM? ? Surface of last scattering M. Tegmark 4 Big. BOSS @NOAO

Big. BOSS: The Stage IV BAO Experiment BOSS Big. BOSS-N JDEM Big. BOSS-N+S Redshift 0. 2<z<0. 7 0. 2<z<3. 5 0. 7<z<2. 0 0. 2<z<3. 5 Sky Coverage 10000 deg 2 14000 deg 2 20000 deg 2 24000 deg 2 Field-of-View 7. 0 deg 2 0. 6 deg 2 7. 0 deg 2 Number of Fibers 1000 4000 Slitless 4000 Angular size of Fibers 2” 1. 5” n/a 1. 5” Wavelength Range 360 -1000 nm 340 -1130 nm 1100– 2000 nm 340 nm– 1130 nm Spectral Resolution 1600 -2600 2300 -6100 DETF Fo. M 57 175 250 286 DETF Fo. M w/Stage 107 240 313 338 III Big. BOSS-North and the full Big. BOSS experiment compared to the current BOSS experiment (under construction) and JDEM (the only other stage-IV BAO project currently proposed). The DETF Fo. Ms include Planck priors or Planck plus Stage III supernova and weak lensing experiments per Fo. MSWG. JDEM (also known as IDECS) Fo. M entries are for JDEM BAO parameters as determined by the JDEM Science Coordination Group, and are consistent with the version submitted to Astro 2010. 5

@ Kitt Peak 4 -m telescope 1. 5 -m f/5 secondary enables 3° FOV 3 -element corrector 4000 fiber positioners on 99 -cm focal plane fiber run (bare fibers) 8 spectrographs BOSS/WFMOSstyle 6

Big. BOSS: The Stage IV BAO Experiment Spectrograph Design Collimator mirror Notional design from JHU Based upon BOSS/WFMOS design Blue “QSO Lyα channel” 34005500 Å at R~4000 e 2 v CCDs Exit fibers on slit-head Blue camera (6 lenses) Blue dewar (2 lenses + CCD) No prisms in Big BOSS Gratings Visible “supernova channel” 5500 - Dichroic (only blue light reflected) 8000 Å at R~3500 Red camera LBNL CCDs (not shown) (6 lenses) Red “galaxy channel” 8000 -11, 300 Å at R~5000 LBNL CCDs + Teledyne Hg. Cd. Te Red dewar (2 lenses + CCD/Hg. Cd. Te) 8

Big. BOSS is Low Risk • Strongest galaxy emission lines are: OII (373 nm doublet), and Ha (658 nm). The OII line flux is well-known even at high redshifts (z=1. 5) from DEEP, COSMOS, 2 d. F. The use of the OII line leads to the simplest, lowest cost system. • We can guarantee a source density. • A targeted survey is efficient, we only spend valuable observing time where we have too, and we can sculpt the distribution to obtain a uniform shot noise limited sample. • Very simple data processing (spectral extraction) developed for BOSS, and no source confusion. • The OII doublet is fully resolved by the Big. BOSS spectrograph and is a unique signature that will not lead to redshift errors due to single line detections. • A dark energy measurement requires 10 -4 calibration bias of the redshifts, which is straightforward in a high resolution spectrograph.

Question #1 Question: “ What fraction of the observing time do you expect to be devoted to the Big. BOSS survey during its operation? What information do you have on the likelihood that the US community will approve of devoting a huge fraction of the 4 m dark time the Big. BOSS science for 2015 -2021? What is the process for approving this allocation of resources? ”

Question #1 First Part: “ What fraction of the observing time do you expect to be devoted to the Big. BOSS survey during its operation? ” • The full Big. BOSS experiment uses 6 years of dark time on a 4 -m telescope to obtain its complete sample of 30 million galaxies to z=2. 0 in the Northern sky. • An additional 4 years of dark time on Southern telescope would complete the 50 million galaxy sample over 24, 000 deg 2. • This time assumes 50% loss of time due to weather, and a 60% observing speed relative to good conditions. • A compelling experiment can be performed with to z=1. 6 with 100 nights/yr (equivalent to the DES utilization of CTIO) for 10 years (North+South) yet still obtain 50 million galaxies yielding a Fo. M about 15% less.

Question #1 Second Part: “What information do you have on the likelihood that the US community will approve of devoting a huge fraction of the 4 m dark time the Big. BOSS science for 2015 -2021? What is the process for approving this allocation of resources? ” • There are seven 4 -m class facilities in the US OIR System: Palomar 5 m, SOAR 4. 2 -m, KPNO 4 -m, CTIO 4 -m, WIYN 3. 5 -m, ARC 3. 5 -m, and Shane 3. 0 -m, and Lowell 4. 2 -m (available in 2 years). • We have “baselined” the Kitt Peak 4 -m and CTIO 4 -m. We are systematically exploring optical designs on the 14 astronomical telescopes in the 3. 5 to 4. 2 -m aperture range. • Of the 5 telescopes studied to date, the Kitt Peak 4 -m, CTIO 4 -m, and Calar Alto 3. 5 -m are suitable to conversion to a 3 -degree field.

Question #1 • NOAO has developed a plan for current and future needs of telescope access on apertures smaller than 6 -m through its Re. STAR committee (Renewing Small Telescopes for Astronomical Research). • This committee’s recommendations call for the specialization of the 2 -4 meter class telescopes: “Specialization will provide a more limited set of observing capabilities on each telescope but should preserve a breadth of capability across the Re. STAR System. ” • The Big. BOSS instrument would be the most ambitious low- and midresolution spectrograph ever fielded on a 4 -m class telescope, with 5000 fibers spanning the full 340 -1150 nm wavelength range.

Question #1 • At the NOAO 4 -m telescopes, the community would have direct access to the capability in four ways. 1. First, through the NOAO TAC the community could directly apply for nights with the instrument. 2. Second, through the NOAO TAC, the community could propose for fibers to be used during normal survey operations. 3. 4. Legacy value of the survey data Participation in the collaboration

Question #1 Third part: “What is the process for approving this allocation of resources? ” • In the process that selected DES, NOAO to issued a call for large proposals. • That process constrained the available dark nights to 100 nights per year. • Two factors could allow that number to grow. First, displacement funding could be provided (to NOAO) to purchase nights on other 4 -m class telescopes. The Re. STAR committee has already recommended that option as a means to obtaining access to new capabilities.

Question #2 Q: “Can you provide a detailed cost estimate, including operations? What is included in the operations cost estimate? Which parts of the path from data to science are not included, and where will that funding come from? ”

Question #2 Costing Methodology • Costs based on the Big. BOSS schedule with a four year development from the project initiation to the start of operations • Big. BOSS relies heavily on heritage borrowed from BOSS — Detailed, as built, expenses for the spectrographs, detectors, electronics, fibers, computing, and operations • Instrument and telescope operations costs provided by KPNO • Though we expect to have significant reuse of the data pipelines from BOSS, we have included significant funds for pipeline development • Include significant costs for data reduction and creation and export of data catalogs to the community • Have not included cost for scientific analyses — Expect this to be borne by our collaborating faculty through their academic support

Question #2 Big. BOSS Total Cost

Question #2 What is included in Ops Cost? • Current operations cost for telescope operators and associated management and administrative support • Additional personnel needed for Instrument support • Data reduction effort estimated at the same level as for the BOSS project, with an additional year of effort before the start of operations • Additional software project management at the level of 0. 5 FTE for four years • Pipeline and data operations funded until the final data release, six months after the end of operations

Question #2 What is not included in Ops Cost • The budget does not include data analysis beyond the reduction of the data to 1 -D spectra, redshifts and object classification — The highly successful SDSS-I, SDSS-II and SDSS-III models will be followed in which the collaboration members carried out these analyses — In practice, member universities have typically committed funds both for the project buy-in and for science postdocs

Question #2 Agency Costs • We have developed a provisional concept for the international collaboration and are focused on adopting a funding model that was used successfully on BOSS • For Big. BOSS our expectation is that DOE along with international partners and University in-kind contributions would be the principal funding mechanism for the instrument — operations costs would be provided by NSF and University funds at our collaborating institutions • We have included the NOAO costs of the telescope operations in the table of costs above — Nominally NOAO would assume these costs — Breakdown is shown in the table above

Question #2 Big. BOSS Cost Descope Option • Same instrumentation except that the NIR detectors are omitted — Saves $9. 7 M in NIR detector cost — Science descoped to cover range of z<1. 6 instead of z<2. 0 — Yields significant reduction in agency funding

Question #3 Q: “What line flux sensitivity do you expect as a function of wavelength in a one-hour exposure (for all redshifts in the survey)? What comoving number density of galaxies as a function of redshift will the survey sample? ” • SFG’s: — The Big. BOSS science goal is to measure redshifts of split [OII] line emission from star-forming galaxies (SFGs) between 0. 7 < z < 2 at a source density of dn/(dz deg 2 )=2000. — Corresponds to a constant comoving number density of 3. 4 x 10− 4 (h/M pc)3 — Single-line minimal detectable line flux (MDLF) of 2. 5 x 10 -17 • LRG’s: — also selected with same number density out to z=1 and will be better BAO tracers since their dark matter halo mass bias is a factor of 2 larger than the SFGs. • QSO’s: — target density will be sparsely sampled in the manner of BOSS, but the 4 m aperture of Big. BOSS will allow us to sample 1 million sightlines from 2 < z < 3. 5.

Question #3 Blue and visible arm

Question #3 red arm

Question #4 Q: “How technically risky is the 3 degree field? If it is not achievable, what is the descope option and how does it modify the science achievements? ”

Question #4 Q: “How technically risky is the 3 degree field? If it is not achievable, what is the descope option and how does it modify the science achievements? ” Element Diam (m) Material Optics Secondary 2. 0 ULE or Zerodur asphere C 1 1. 5 Fused silica sphere Surf 1 Surf 2 C 2 Surf 1 sphere 1. 2 Fused silica Surf 2 C 3 Surf 1 Surf 2 sphere asphere 1. 3 Fused silica asphere

Question #4 • To accommodate a 5, 000 automatic fiber positioner system our conceptual design requires a 2. 5º field-of-view including space for the telescope guiding system. • The 3º field of view (or 2. 5º) has not proved to be a technical challenge with constraints such as optics, telescope mechanics, baffling, or vignetting. • Mayall is a fairly slow Ritchey Chretien telescope, which reduces the optical challenge of the design, and allows the 3º field. • The secondary mirror magnifies the prime focus, and the corrector assembly is not required to do so. Chromatic aberration is not a strong driver in the design of the corrector. All three elements may be made from fused silica. No opto-mechanically challenging materials (such as calcium fluoride) are necessary for color correction.

Question #4 • Manufacturing feasibility of the secondary mirror, as well as the corrector elements, were verified by the University of Arizona College of Optical Sciences. Using profilometry and interferometry, all elements are less challenging than optics made by that facility in the past • Another possible descope is reduction in the size of the secondary mirror. The 2 -m mirror is currently properly sized with a 37% obscuration and no field-dependent vignetting. If the secondary mirror size is reduced from 2 -m to 1. 5 -m diameter, the vignetting becomes field-dependent with the same 37% obscuration in the center, increasing to a 55% effective obscuration at the edges.

Question #5 Q: “Please present a detailed estimate and justification of the accuracy of photometric redshifts in the 1. 3 < z < 2 range and the predicted OII line flux, and how it translates to object pre-selection efficiency. ”

Question #5

Big. BOSS: The Stage IV BAO Experiment Corrector + Fiber Positioners Updates since white paper on April 1 st Corrected field difficult to baffle beyond 2. 5 deg diameter (not 3. 0 deg) Fiber positioner design can be as small as 1. 0 cm center-to-center (1. 1 cm is “easy” to build) 2. 5 deg field (82. 5 cm) with 1. 25 cm positioners → 4000 positioners possible 2. 5 deg field (82. 5 cm) with 1. 10 cm positioners → 5100 positioners possible 2. 5 deg field (82. 5 cm) with 1. 05 cm positioners → 5600 positioners possible 2. 5 deg field (82. 5 cm) with 1. 00 cm positioners → 6170 positioners possible Do we want 10 -11 spectrographs with 500 fibers each? 32

Big. BOSS: The Stage IV BAO Experiment Conclusions • A “Stage-IV” dark energy scientific program from the ground – 0. 2 < z < 2. 0, and up to 3. 0 with QSO’s • • Equivalent to JDEM: large sky coverage, redshift, and resolution Uses sample to find visible emission-line sources (OII, OIII) – Up to 50 million targeted emission-galaxies in 10 years • • • SDSS BAO discovery was 60, 000 galaxies BOSS will have 1, 500, 000 galaxies, 0. 3 < z < 0. 7 JDEM uses a blind search and finds more galaxies, but not better figure-of-mert • Complementary to large imaging surveys (DES, LSST) – – Adds spectroscopic capability, eg. for SNe follow-up Calibrates LSST photo-z’s for WL • BOSS lays groundwork for Big. BOSS – – Reuse spectrograph design Reuse pipeline • SNAP lays groundwork for Big. BOSS – – Reuse SNAP detector & electronics development Mini focal planes (8) with 2 x. CCD’s and 2 x. Hg. Cd. Te’s each • Requires only 4 -m telescope time – – North: Kitt Peak (4 m) South: CTIO (4 m) 33

Big. BOSS: The Stage IV BAO Experiment Science Reach vs. JDEM Redshift BOSS (Stage III) 0. 2<z<0. 7 Big. BOSS-North JDEM (Stage IV) 0. 2<z<3. 5 0. 7<z<2. 0 Big. BOSS-N+S (Stage IV) 0. 2<z<3. 5 Sky Coverage 10000 deg 2 14000 deg 2 20000 deg 2 24000 deg 2 Field-of-View 7. 0 deg 2 0. 6 deg 2 7. 0 deg 2 Number of Fibers 1000 4000 Slitless 4000 Angular size of Fibers 2” 1. 5” n/a 1. 5” Wavelength Range 360 -1000 nm 340 -1130 nm 1100– 2000 nm 340 nm– 1130 nm Spectral Resolution 1600 -2600 2300 -6100 DETF Fo. M 57 175 250 286 240 313 338 DETF Fo. M w/Stage III 107 Big. BOSS-North and the full Big. BOSS experiment compared to the current BOSS experiment (under construction) and JDEM (the only other stage-IV BAO project currently proposed). The DETF Fo. Ms include Planck priors or Planck plus Stage III supernova and weak lensing experiments per Fo. MSWG. JDEM (also known as IDECS) Fo. M entries are for JDEM BAO parameters as determined by the JDEM Science Coordination Group, and are consistent with the version submitted to Astro 2010. 34

Big. BOSS: The Stage IV BAO Experiment Science Reach vs. JDEM Figure 1 a: Distance accuracies in Δz=0. 1 bins for Big. BOSS (red) and JDEM (blue) normalized to the cosmic variance limits. These forecasts were based on the Seo & Eisenstein (2007) Fisher matrix formalism and assume a 50% reconstruction of the acoustic feature. Figure 1 b: The inverse variance on the first 30 principal components of the evolution of the dark energy, as defined by the Figure of Merit Science Working Group (Fo. MSWG). The variances have been normalized to the pre-JDEM Stage III forecasts made by the Fo. MSWG.

Big. BOSS: The Stage IV BAO Experiment R&D Funding + WBS FY 09 FY 10 FY 11 (Q 1 only) Concept Design 260 375 100 Spectrograph Prototype 370 850 250 Focal Plane Prototype 0 465 120 Fiber Positioner 910 230 2600 700 620 Total 1250 Table 4. Big. BOSS R&D Funding Plan Above R&D activities supported by LBNL engineering and technical labor CD-0 assumed FY 11 Q 1