Advancing Line Detection Rates for Submm Spectroscopic Surveys: Insights from BLISS/X-Spec Teams

1. Line Detection Rates for Next Generation IR/Submm Spectroscopic Surveys Eric J. Murphy BLISS/X-Spec Science teams

2. Meet Dylan

3. Why mid-/far-IR Spectroscopy: BLISS? Circinus Galaxy ISO SWS+LWS • Many bright atomic fine-structure + access to H2 lines and PAH features • Cosmic Star Formation History • SFR, gas density, etc. • Cosmic Rise of Heavy Elements • PAH emission at z~6 • Black Hole Birth • Extreme species e.g., [NeV] and high-J CO lines • Gas in Forming Planetary Systems • Gas cools through FIR atomic O and C transitions -- gas-disk lifetimes BLISS band at z=2 Detectable at z~8-10 with BLISS! BLISS band at z=4 Egami et al 2006, Spitzer IRS

4. Why Submm Spectroscopy: X-Spec • Many bright atomic fine-structure and molecular rotational transitions detectable. • Spectroscopic redshifts and interstellar gas conditions from galaxies in the early universe (i.e., z > 6) • e.g., z=6.42 Walter et al. 2009). • Much faster (~30x) than ALMA for survey work! Z-spec Spectra from Bradford+ (2011/2010) z=2.56

5. BLISS/SPICA Specs.

6. X-Spec/CCAT Specs. • First Light Instrument for CCAT • Primary Mirror: 25m • chopping Secondary? B1: 575 – 945μm B2: 965 – 1535μm • MDLF (3σ, 1hr): 0.72 – 7.2 x 10-20 W m-2 • SuperSpec chip with MKIDs • Simpler readout architecture than TES bolometers • 100-1000 detectors per wire vs. 5-20 • Two potential implementations: • Direct Imaging Spectrometer • Steered Beam Multi-object Spectrometer

7. BLISS/X-Spec Sensitivities Put into Context Matt Bradford

8. Is confusion going to be an issue? Constructing Line Count Models • Galaxy Number Count Models: N(z, LIR) • E.g., Chary & Pope (2012) • Uses the deepest 24μm data & Spec-z’s from GOODS • Constrained by the Cosmic IR Background • Consistent with Herschel observations • Does not take clustering into account • Empirical Line Prescriptions (31 lines included) • 15 mid-/far-IR lines: Spinoglio et al. (2012) • 16 CO & other submm line: Visbal & Loeb (2010)

9. Integral Line Counts: BLISS/SPICA 1 hr • More lines at longer wavelengths • Naturally, since beam increases w/ λ • Also, more species available (8 to 16 between • 50 and 350μm) • At 250μm, one line per75beams detected • (5σ) with > 1.5 x 10-20 W m-2

10. Integral Line Counts: X-Spec/CCAT 70 hr 1 hr • More lines at longer wavelengths • Naturally, since beam increases w/ λ • Number of available species ~17 at all λ • At 1.3mm, 70x 1hr pointings (300 beams/pt) to • detect a single line (5σ) with >1.5 x 10-20 W m-2

11. So, even with BLISS, confusion is not an issue… But can we extract lines from spectra? Intrinsic Source Spectrum 10 Examples Blank Field Spectrum Assuming two off nod positions • Confused Spectra • Monte-Carlo number counts between NGC520, Mrk231, and Arp220 • Mock Observation of NGC520 at z=4 scaled to LIR 2x1012L

12. Findings & Remaining Issues • BLISS: • Line Confusion does not seem to be an issue • How well can identify & recover line fluxes in spectra? • Quantify how the continuum shapes of sources are affected by intervening sources (needed for Md, Td, Ld). • Is R=400 too course spectral resolution? • X-Spec • Many pointings to measure1 line even with 300 beams • Steerable system or a integral field unit?

14. BLISS Specs.

15. X-Spec/CCAT Specs.