Observing Cosmic Dawn with the LWA1Jake Hartman, Judd Bowman, and Greg Taylor

on behalf of the LWA collaboration

Current status of the LWA• 10–88 MHz aperture synthesis radio telescope• First station (LWA1, at the EVLA) is nearly operational• Leases obtained for LWA2 and LWA3 sites

(baseline lengths 19 km, 35 km, and 43 km)• Goal is 53 stations with baselines up to 400 km for resolutions

of 2” at 80 MHz (8” at 20 MHz) with ~mJy sensitivity

LWA1 capabilitiesFrequency range 10–88 MHzNumber of beams 4Number of tunings 2 per beamBandwidth per tuning 0.2–16 MSPSPolarization isolation ≥ 20 dBBaseline lengths 5–110 mBeam width 2° × (80 MHz / ν)Sky coverage elevation ≥ 20°Thermal noise 20 mJy (1 hr, 8 MHz BW)

102

103

104

105

106

107

108

109

10 20 30 40 50 60 70 80 90

— Sky— Terminated

Frequency (MHz)

Tem

pera

ture

(K)

Good RFI environment• Sky-noise dominated by a

factor of ≥4 in 25–87 MHz• Less than ~10% loss to RFI• Deep 1 hr integration shows

no systematic noise sources (2.5 MHz BW at 72 MHz)

Good RFI environment• Sky-noise dominated by a

factor of ≥4 in 25–88 MHz• Less than ~10% loss to RFI• Deep 1 hr integration shows

no systematic noise sources (2.5 MHz BW at 72 MHz)

100 101 102 10310–6

10–5

10–4

10–3

10–2

10–1

Integration time (s)

Varia

nce

(dB)

Measured noiset –1/2

✴

—

The global 21 cm signatureA. Falling gas density decouples TS from Tgas

B. The first stars form. Ly alpha photons recouple TS to Tgas via the Wouthuysen-Field effect.

C. X-rays from compact objects left behind by deaths of the first stars begin heating Tgas.

D. Reionization begins, reducing the HI fraction.

E. Reionization is complete.

(Pritchard & Loeb 2010)

Requires 1 part in ~105 relative spectral calibration!

(Oliveira-Costa et al. 2008)50 MHz model of northern sky

Cosmic Dawn with the LWA1• Observations use all four beamformers to make two effective

beams, each covering 32–88 MHz• Science beam targets a relatively cold region of the sky• Calibrator beam targets a bright, smooth spectrum source• Beams are large enough to average over angular variations

Science beamAvg. 4200 K

Calibrator beamCyg A: 77000 K

Bandpass calibration• Each beam’s signal = G(ν) × [Tfore(ν) + ΔT21(ν)]• Simultaneous beams mean same signal path, so shared G(ν)• Foreground terms must be modeled but should be smooth,

with 0–1 inflection points in our bandpass (vs. 3–4 inflection points for the expected 21 cm signal)

• Uncertainties due to bandpass calibration should be ~10 mK with 100 hr integration, enough to constrain z ~ 20 feature

• Measurements can be repeated

Frequency dependence of sidelobes• Must prevent frequency-dependent variations in the beams’

sidelobes from coupling with foreground structure• Defocusing the science beam

+ Averages over more foreground; can lower sidelobe power+ Requires different antenna weighting between beams

• Steering sidelobes away from bright sources+ Requires excellent model of the electrodynamic profile of station

• Shimmering sidelobes• Optimal beamforming: accounting for mutual coupling, etc.

+ Requires excellent model of the electrodynamic profile of station

LEDA• “Large-aperture Experiment to detect the Dark Age”

(ATI proposal; PI: L. Greenhill)• Goal: observe global 21 cm signal at 38–88 MHz (z = 15–36)• New hardware:

+ Wide-bandwidth correlator backend based on CASPER ROACH II boards (F engine) and GPUs (X engine)

+ Addition of a switchable calibrated noise source to the front end electronics of a subset of the antennas

• Initial tests to be performed at LWA1; full installation would likely be at a future LWA station

• Both approaches complementary to DARE (next talk)

Observations with the full LWA• Full 53 stations would provide

arcsecond resolution• Imaging of the spatial variability

in the heating by first stars, black holes

• Currently unfunded(cost per station: ≈$1.5M)

• Would enable a great range of exploration: transient LF sky, extrasolar planets, pulsars, radio galaxies, ionosphere, solar science and space weather, …

10 km

Summary• LWA1 offers a novel method to measure or constrain the all-

sky 21 cm signal using large beams• Bandpass calibration accomplished by comparing science

and calibrator beams• Will develop advanced beamforming techniques• Could begin measuring or constraining the early universe with

~100 hours of integration• LEDA offers a complementary (lower risk, higher cost)

approach