Transcript chime
C anadian H ydrogen I ntensity M apping E xperiment z 0 Age of Universe Big Bang CMB CHIME ∞ 9 5 3 2 1 13.7Gy SDSS WiggleZ Reionization Keith Vanderlinde CIfAR C&G 2012 AGM C anadian H ydrogen I ntensity M apping E xperiment Outline • Cosmo Context • HI Mapping of BAO • FFT Telescopes • CHIME Keith Vanderlinde CIfAR C&G 2012 AGM Composition of the Universe -4 -3 Log (a) -2 -1 1 0 Frieman et al. 2008 ? w = -1 ± 0.2 • Can study via growth of space. • Need standard ruler over wide z range! Composition Images from NASA / WMAP Science Team • Photons • Neutrinos • Dark Matter • Electrons + Baryons Hu & White, Scientific American, 2004 The Primordial Plasma Baryon Acoustic Oscillations • random background of primordial fluctuations • each fluctuation emits a pressure wave • at decoupling, the size of this wave is frozen Measured in the CMB Cold Stack (Eisenstein) Hot Stack Komatsu et al. 2010 Post-Decoupling Radial Profile Eisenstein BAO peak BAO are “frozen in,” ~150 co-moving Mpc. BAO Observations Detected in galaxy correlations! Buetler et al, 2011 Eisenstein et al, 2005 SDSS 6dFGS Blake et al, 2011 WiggleZ mh2 = 0.12, mh2 = 0.024 mh2 = 0.13, mh2 = 0.024 mh2 = 0.14, mh2 = 0.024 mh2 = 0.105, mh2 = 0.0 Sanchez et al, 2012 BOSS-CMASS BAO as a standard ruler Anderson et al. 2012 BAO “Size” at z WMAP ... ... 1100 How do we fill this in? Optical Maps of Structure Get spectroscopic redshifts on lots of galaxies. Works! BAO Scale But: • expensive • hard at z>1 Blanton et al. (2003) Sloan Digital Sky Survey (SDSS) Hydrogen Intensity Mapping (See Ue-Li’s talk!) For BAO, no need to resolve galaxies individually. Use 21cm HI line. For 15 co-moving Mpc resolution at z~1-2 (~20’), need: a) 100m telescope + decades of survey b) Interferometer + HUGE correlator c) FFT telescope BAO Scale Unresolved galaxies, resolved BAO. FFT Telescope +++++++ +++++++ +++++++ +++++++ +++++++ +++++++ +++++++ Sampled Ground Ground E Sky Measured Sky Haslam et al. 1982 Normal Telescope noise ~ Tsky + Tamp x Use a mirror (or lens) to focus light onto a detector. Phased Array Interferometer noise ~ Tsky + Tamp sqrt(Namp) x x x x x x x x x Record signals at the dish, add delay to mimic reflection, sum recorded signals. Phased Array Interferometer noise ~ Tsky + Tamp sqrt(Namp) x x x x x x x x x Throw away the dish. Save a big pile of money. “Pointing” a Phased Array noise ~ Tsky + Tamp sqrt(Namp) x x x x x x x x x Add delays with cabling to mimic pointing. Spend your pile of money on a computer, do this digitally. Beam-forming Telescopes Record signals on the ground, digitally delay and add. noise ~ Tsky + Tamp sqrt(Nant) -- per beam! -- x x x x x x x x x N beams, without aliasing, for spacing < /2. Mapping speed ~ Nant x single-dish. (N2 for Tsky >> Tamp) Cylinder Telescopes Hybrid: 1D Dish + 1D FFT Single Antenna Long Narrow Beam Cylinder Telescopes Hybrid: 1D Dish + 1D FFT N “Formed Beams” N antennas Canadian Hydrogen Intensity Mapping Experiment 100m 100m • 400-800MHz band 21cm from z ~ 0.8 - 2.5 • Resolution: 1MHz, 13-26’ <10% BAO scale •Drift scan, no moving parts 20,000 deg2 coverage Readout & Correlator 1 1) Sample & channelize • 1 ADC per antenna • 8-bit @ 800MSPS 2 2) Correlate Record Data • 1 node per freq band • N2 is tricky: Pflops of processing, Exabytes of storage (!!) • Do FFT beamforming, plus partial N2 to calibrate CHIME Sensitivity to BAO spectrum w/ BAO spectrum w/o BAO CHIME Cosmology Forecast (1+z)3 ^ (Preliminary!) Pulsars & Radio Transients Every day, CHIME observes: • every point in the northern celestial hemisphere • across an octave in frequency (400-800MHz) • for between 5 minutes and hours (depending on Decl.) • with s timing resolution & accuracy Can’t actually process every point at that rate. Tee data off to monitor known locations, systematically search for new sources. 0 Mapping the Observable Universe Age of Universe Big Bang CMB CHIME (400 co-moving Gpc3) z ∞ CHIME will: • survey BAO • measure the growth of space • from 0.8<z<2.5 • over a volume of ~400 co-moving Gpc3 9 5 Reionization 3 2 1 13.7Gy SDSS WiggleZ CHIME Survey Volume For cosmic variance limited survey, sensitivity ~ volume. CHIME’s sensitivity to D.E. within ~2x of c.v. limited. The End