C-BASS C-Band All-Sky Survey Tim Pearson (Caltech) 2009 July 2 - PowerPoint PPT Presentation

C-BASS C-Band All-Sky Survey Tim Pearson (Caltech) 2009 July 2

Summary • Image the whole sky at 5 GHz (“C band”). • In both brightness and polarization . • Broad-band (1 GHz) correlation polarimeter and correlation radiometer. • Two telescopes: one in California, and another in South Africa. • FWHM 0.85° – similar to Haslam-408, WMAP. • rms noise < 0.1 mK in I, Q, U • Completion in 2011(?) to support Planck analysis. • Northern survey: start 2009 • Southern survey: start 2010 2 Tim Pearson 2009 Jul 2

Motivation • CMB Task Force recommendations (2005): • “A systematic program to characterize astrophysical foregrounds, especially from the galaxy, over a wide range of frequencies.” • “Continued support for ground-based efforts to produce 3–15 GHz large-scale maps of the polarized Galactic foreground.” 3 Tim Pearson 2009 Jul 2

Science Goals • Survey of diffuse Galactic emission at a frequency low enough to be dominated by synchrotron radiation but high enough to be uncorrupted by Faraday rotation effects. • Enable accurate subtraction of foreground contaminating signals from higher-frequency CMB polarization sky surveys, including WMAP and Planck . • Major resource for studying the interstellar medium and magnetic field of the Galaxy. 4 Tim Pearson 2009 Jul 2

CMB Foregrounds • Low-frequency foregrounds: • Synchrotron • Free-free • “Spinning dust” • C-BASS at 5 GHz is dominated by synchrotron • A polarized synchrotron template ? – input for foreground modeling (e.g., spatial variation of spectral index and curvature). • 5 GHz is lowest frequency where Faraday rotation is negligible, < 1° except in Galactic plane 5 Tim Pearson 2009 Jul 2

What does C-BASS add? • Compare Planck alone to Planck + C-BASS • Single pixel analysis (1000 realizations), simple Galactic model • Clive Dickinson FGFIT (MCMC parametric fitting) • Typical high-latitude pixel (2° beam): • Spectral index bias reduced: Stokes I: − 0.14 → 0.015 • Stokes Q,U: − 0.16 → 0.03 • • 70 GHz synchrotron amplitude error reduced: Stokes I: σ : 0.9 μ K → 0.3 μ K (SNR: 3.5 → 12) • Stokes Q,U: σ : 0.3 μ K → 0.045 μ K (SNR: 1 → 7) • • 70 GHz synch. amplitude bias reduced: Stokes I: 0.9 μ K → 0.15 μ K • Stokes Q,U: 0.015 μ K → 0.003 μ K • • 5–7 times reduction in synchrotron residuals in the CMB band! 6 Tim Pearson 2009 Jul 2

WMAP5 23 GHz map of polarized intensity (color) and direction (vectors) from WMAP (Hinshaw et al. 2008). DRAO 1.4 GHz Polarized intensity (Wolleben et al. 2006 A&A 448, 411)

Predictions I Q U (Planck Sky Model) 8 Tim Pearson 2009 Jul 2

Survey Requirements • Sensitivity: In order to subtract the polarized Galactic foregrounds to below the sensitivity levels of Planck requires an rms noise level of < 100 µK per pixel. Our goal is to produce a substantially lower noise level and reduce systematic errors to well below 5% level. • Resolution: To detect the B-mode peak at l ~ 90 we need measurements up to l ≈ 150, which fixes the resolution of the survey to about 1°. • Frequency: High enough to avoid Faraday rotation, low enough to maximize sensitivity to synchrotron. • Bandwidth: Limited by manmade interference (RFI). 9 Tim Pearson 2009 Jul 2

Collaboration • Caltech/JPL/OVRO: Dayton Jones, Russ Keeney, Charles Lawrence, Erik Leitch, Stephen Muchovej, Tim Pearson, Tony Readhead, Graça Rocha, Matthew Stevenson. • Northern survey, OVRO antenna, backend and data acquisition. Supported by NSF. • Oxford University: Christian Holler, Jaya John John, Mike Jones, Oliver King , Angela Taylor. • Feed optics, receiver and polarimeter, cold loads. • Manchester University: Rod Davies, Richard Davis, Clive Dickinson,Tess Jaffe, Paddy Leahy, Stuart Lowe, Neil Roddis, Althea Wilkinson, Peter Wilkinson. • Low-noise amplifiers. • Rhodes University / HartRAO: Roy Booth, Charles Copley, Justin Jonas. • Southern survey. 10 Tim Pearson 2009 Jul 2

Antennas 6.1m antenna at OVRO, California 7.1m antenna in South Africa (donated by JPL) (moved to a site in the Karoo) 11 Tim Pearson 2009 Jul 2

Systematics • Rapid scanning of the telescope in AZ to reduce 1/ f systematics. • Highly-redundant coverage at a variety of scan crossing angles. • Optical layout and feed-horn have optimized for minimal sidelobes and cross-polarization. No subreflector support legs. • Absorbing tunnels reduce the sidelobes to more than 40 dB below the main beam, while contributing ~ 0.8 K to the system temperature. • Ground screens to shield the receiver from polarized ground-reflected radiation and RFI. 13 Tim Pearson 2009 Jul 2

Scanning Strategy • Constant elevation scanning: constant ground and atmosphere loading • Many scan crossing angles at each pixel to reduce systematics • Scan speed ~ 6°/sec • Many orientations of the polarimeter at each pixel • 50% of each night through the pole (baseline reference) • 50% through pole + 22° 14 Tim Pearson 2009 Jul 2

Scanning Strategy 15 Tim Pearson 2009 Jul 2

Antenas and Optics No feed support legs, absorbing tunnels 16 Tim Pearson 2009 Jul 2

Calculated beam patterns ---- with absorber ----E-field ---- without absorber ----Cross polar Credit: C.M. Holler (Oxford) 17 Tim Pearson 2009 Jul 2

Correlation Receiver ∆ T = ∆ G T sys G ∆ T = ∆ G T A − T ref T sys G T sys 19 Tim Pearson 2009 Jul 2

Receiver layout: I, Q, U T sys < 20 K 4.5 to 5.5 GHz U ∝ � E L E R e i π / 2 � Q ∝ � E L E R � σ Q,U < 0.1 mK 20 Tim Pearson 2009 Jul 2

OMT Cross-pol < -58 dB over 40% BW Return loss ~ -20 dB Very compact, easy to cool to 4 K (Grimes et al. 2007) 21 Tim Pearson 2009 Jul 2

Polarimeter Components Phase switch 26 dB amplifier 180° hybrid Bandpass filter + eMerlin C-band LNAs Oliver King (Oxford) 22 Tim Pearson 2009 Jul 2

Digital Readout • Based on particle physics readout boards developed at Oxford • On-board FPGAs perform subtraction demodulation, integration 23 Tim Pearson 2009 Jul 2

RFI 4.5 – 5.5 GHz Passive Microwave Imagery at 6.9 GHz from AMSR-E on the NASA EOS Aqua platform. 24 Tim Pearson 2009 Jul 2

RFI Monitor • Heterodyne design based around the CASPER iADC and iBOB boards. • Entire 1GHz band is Nyquist sampled, with all DSP occurring on the FPGA. • 512 element spectrometer (2MHz resolution) is implemented. • Can detect horizontal RFI below the C-BASS detection threshold, allowing for unambiguous flagging of corrupted measurements. 25 Tim Pearson 2009 Jul 2

Schedule • OVRO telescope commissioning (Apr-Jul 2009) • Receiver installation and commissioning (Jul-Sep) • Northern survey (1 year) • nights only, allowing 50% efficiency • Build second receiver (multichannel?) • in collaboration with King Abdulaziz City for Science and Technology (KACST), Saudi Arabia; Yaser Hafez • Southern survey (1 year, starting Feb 2010) • Data release in 2011-12? 26 Tim Pearson 2009 Jul 2