Observing Cosmic Dawn with the LWA-1 PIs: Judd Bowman (ASU), Greg Taylor (UNM) Jake Hartman (JPL) Jayce Dowell, Joe Craig (UNM) Steve Ellingson (Virginia Tech) Jackie Monkiewicz Arizona State University
Jan 30, 2016
Observing Cosmic Dawn with the LWA-1
PIs: Judd Bowman (ASU), Greg Taylor (UNM)
Jake Hartman (JPL) Jayce Dowell, Joe Craig (UNM)Steve Ellingson (Virginia Tech)
Jackie Monkiewicz Arizona State University
The “Dark Ages” and Cosmic Dawn
2/20
Dark Ages: z = 1,100 to z~ 40 • matter-dominated• H & He are neutral•1st structures collapsing
Cosmic Dawn: z = 40-20•1st stars & galaxies•1st QSOs?•Early heating, reionization of small bubbles
Cosmic Dawn project purpose:
3/20
Detect/constrain signal of 1st generation of stars in 21-cm absorption of hydrogen at z ~ 30
REQUIRES:
1.Low frequency experiment, 10-100 MHz LWA-1
2.Long Integration time
3.Very accurate bandpass calibration
4. Novel beamforming techniques
Cosmic Dawn in 21 cm:
(Furlanetto 2006, Pritchard & Loeb 2010) 4/20
Tb = Ts – Tcmb
Seen against CMB:
Thermal History of IGM
Cosmic Dawn in 21 cm:
• 1st stars create absorption trough
• Additional heating sources mitigate trough
(Furlanetto 2006, Pritchard & Loeb 2010) 5
Tb = Ts - Tcmb
Observing strategy:
REQUIREMENTS:Very good bandpass calibration!
• Looking for broad, shallow absorption trough• need > 104 S/N in any spectral channel
But only need ~10% accuracy in absolute power…
6/20
A
E
D
B
C
20 40 60 80 100 120 140 160 180
ν (MHz)
ΔT21
(m
K)
+50
0
−50
−100 (Prit
char
d &
Loe
b 20
10)
Observing strategy:
STRATEGY:Simultaneously observe bright calibrators & dark (low Tsys) science field
• 2 x 19.6 MHz beams on bright calibrator• 2 x 19.6 MHz beams on science field• 520 hours on-sky
6/20
Observing Strategy -- COMPLICATION:
• Frequency variation of beam shape couples of foreground structure to sidelobes
mistake sources drifting through sidelobes for 21-cm spectral features?
8/20
1.0
0.8
0.6
0.4
0.2
0.0−15 −10 −5 0 5 10 15
Offset (degrees)
Rel
ativ
e ga
in
74 MHz
38 MHz
Novel Beamforming Strategies:Mitigate potential foreground-frequency coupling of sidelobes:
9/20
1. Defocusing (e.g. gaussian smoothing)
2. Sidelobe steering
3. Nulling
4. Sidelobe shimmering
5. “Optimized” beam-forming(account for mutual coupling of antennas)
Work to date:
10/20
• Learning the LWA Software Library! (and PYTHON in general)
• Phase-and-Sum Beamforming with TBN
• Raster Mapping of TBN Beam (pseudo-beams)
Phase-and-Sum Beamforming:
11/20
Phase-and-Sum Beamforming:
12/20
Find bursting Sun produces much better coefficientsthan Cyg A --- not surprising?
Raster Mapping:
13/20
Use bright source in TBN data to map structure of sidelobes --- “Pseudo-beam”
Raster Mapping – Variation with elevation
14/20
Cyg A:
Transit
EL = 83 deg
-1 hour
EL = 76 deg
-2 hours
EL = 65 deg
15/20
Cas A NCP
@ Cyg A transit
EL = 46 deg
-2 hours before Cyg A transit
EL = 34 deg
@ Cyg A transit
EL = 34 deg
… What is going on in the North/Northeast during the Cyg A transit on Sept 21, 2011??
16
PASI started recording Sept 23, 2011:
http://www.phys.unm.edu/~lwa/lwatv/55827.mov
Pseudo-beam Maps over full frequency range:
17/20
Acquired TBN observations of 4 frequency groups:
87 MHz80 MHz73 MHz
71 MHz64 MHz57 MHz
…corresponding to 4 DRX tunings for main Cosmic Dawn observations
55 MHz48 MHz41 MHz
39 MHz32 MHz25 MHz
Beam 1Tuning 1
Beam 1Tuning 2
Beam 2Tuning 1
Beam 2Tuning 2
Test our Beamforming Strategies:
18/17
1. Defocusing (e.g. gaussian smoothing)
2. Sidelobe steering
3. Nulling
4. Sidelobe shimmering
5. “Optimized” beam-forming(account for mutual coupling of antennas)
Which is the “quickest and dirtiest”?
Test our Beamforming Strategies:
19/17
1. Defocusing (e.g. gaussian smoothing)
2. Sidelobe steering
3. Nulling
4. Sidelobe shimmering
5. “Optimized” beam-forming(account for mutual coupling of antennas)
Which is the “quickest and dirtiest”?
Pseudo-beam Maps over full frequency range:
20/20
Apply some of our novel beam-forming strategies
Defocusing is simplest, fastest apply Gaussian to antenna gains
\
Acquire raster maps of customized DRX beams,compare with TBN predictions, confirm shape
LWA-OCD Project Outputs:
21/20
Beamforming: Detailed measurements of beamStrategies for custom beamforming
Deep integrations:Very high S/N spectra of bright calibratorsVery high S/N spectra of diffuse Galaxy
(including high-level H recombination lines)• Serendipitous radio transients
Lots of opportunity for RFI mitigation!
Detection/contraints on First Light absorption trough.
END
RFI environment at LWA:
22
The “Dark Ages” and Cosmic Dawn
23/17
z = redshift (decreases with increasing time)
z = ∞ z = 0
The “Dark Ages” and Cosmic Dawn
24/17
Wouthuysen-Field Effect:
25
Phase-and-sum Beamforming:
26/17
•Use bright source to back out coefficients•Only works for narrow (< 10 KHz bandwidth)
insensitive to 2 offsets
Delay-and-sum Beamforming:
•Do phase-and-sum over full LWA frequency range•Solve for true delays for each antenna
System noise for LWA:
27From Pihlstrom, 2012, internal memo
Acronyms:
28