The 21cm line from neutral hydrogen can be used to study cosmology during the first billion years of the universe. This includes the Dark Ages when no structures formed, the Cosmic Dawn when the first luminous objects formed, and the Epoch of Reionization when these objects reionized the intergalactic medium. Current and future 21cm experiments like LOFAR, MWA, PAPER, and HERA aim to detect the signal from these eras but face challenges in calibrating the instruments and subtracting bright foreground sources. Some progress has been made in placing upper limits on the signal and constraining the heating of the intergalactic medium by X-rays, but a clear detection of the signal is still needed

1. Cosmology with the 21cm line
Gianni Bernardi
SKA SA
(RARG: O. Smirnov, G. Bernardi, T. Grobler, C. Tasse)
Collaborators: (LOFAR-EoR, MWA-EoR, PAPER?)
AIMS, August 14th 2013

2. The 21cm line is ideal to study the first billion years
Dark Ages: no structures were
formed, primordial
fluctuations are imprinted in
the HI gas
Cosmic Dawn: first luminous
structures (Pop III stars?
Micro quasars?) are formed in
the dark matter halos
Reionization (EoR): luminous
structures (galaxies, AGNs) re-
ionize the IGM

3. mK
21 cm line cosmology

4. Courtesy A. Meisinger
Evolution of fluctuations

5. Observational specs
for 21cm line experiments:
Frequency coverage:
30-200 MHz (6 < z < 35)
Angular resolution:
fluctuations  5 < θ < 30 arcmin  you need a radio interferometer
imaging  up to < 1 arcmin  you need a radio interferometer
Sensitivity:
mK sensitivity is required to constrain most of the HI models
(The VLA @ 74 MHz has an rms sensitivity of 26 K (1 hour))
Challenges:
- correction of ionospheric distortions
- calibration of time and frequency variable telescope response (beam)
- subtraction of bright foregrounds (and their coupling with the instrumental response)

6. Interferometry in 1 slide

7. 2 Antennas
8
Image formation with N antennas

8. 3 Antennas
9
Image formation with N antennas

9. 4 Antennas
10
Image formation with N antennas

10. 5 Antennas
11
Image formation with N antennas

11. 6 Antennas
12
Image formation with N antennas

12. 7 Antennas
13
Image formation with N antennas

13. 8 Antennas
14
Image formation with N antennas

14. 8 Antennas x 6 samples
15
Image formation with N antennas

15. 8 Antennas x 30 samples
16
Image formation with N antennas

16. 8 Antennas x 60 samples
17
Image formation with N antennas

17. 8 Antennas x 120 samples
18
Image formation with N antennas

18. 8 Antennas x 240 samples
19
Image formation with N antennas

19. 8 Antennas x 480 samples
20
Image formation with N antennas

20. 21

21. We live in the era of exploration: current and future 21 cm
experiments
GMRT
LOFAR
PAPER
MWA
HERA - SKA

22. GB et al. 2009
~2.3 arcmin resolution
frequency: ~150 MHz
peak flux ~ 2.8 Jy
conversion factor:
1mJy/beam=4 K
noise: 0.75 mJy/beam
The key point to detect the 21cm signal is how well
foregrounds can be removed! What do foregrounds look like?

23. GB et al. 2009
Statistical properties of foregrounds
* 2noise
180
Y
l b
C X l X l b
N N
 

 


Power law behavior with best fit
amplitude
A400= (0.0019 0.0003) K2,
and best fit slope βI = -2.2 0.3:
diffuse Galactic emission
Flat power spectrum:
residual point sources
Power law behavior with best fit
amplitude A700= (90 7) (mK)2
and best fit slope βIp = -1.52 0.16:
diffuse polarized Galactic
emission

24. Statistical properties of foregrounds: the “wedge”
Pober et al. 2013

25. Do we know how to subtract foregrounds? How well?
• Subtraction of Galactic diffuse emission and extragalactic radio sources:
they are supposed to have smooth spectra compared to the 21 cm signal;
Bowman, Morales & Hewitt 2009

26. Do we know how to subtract foregrounds? How well?
• Subtraction of Galactic diffuse emission and extragalactic radio sources:
they are supposed to have smooth spectra compared to the 21 cm signal;
EoR + FG + noise
Eor + noise
EoR ~ 5 mK
FG ~ 2 K
noise ~ 50 mK
How well does it work on data?
Jelic, .., GB, et al. 2008

27. • An interferometer never samples all the Fourier modes  PSF sidelobes corruption
(k┴,k║);
• Instrumental frequency response corrupts the foreground frequency smoothness (k║);
• Telescope beams change with frequency and pointing direction (dipoles do not track
the sky) and they can be different from each other (k┴,k║);
• The ionosphere is no longer transparent (time and frequency dependent distortion &
refraction) (k┴,k║);
• RFI corrupts the sky signal (mostly ,k║);
• Real foreground polarized signal can leak into total intensity due to polarized beams
a/o imperfect polarization calibration (k║);
The point is that instrument calibration and foregrounds are
coupled  foreground properties are corrupted by the
instrumental response

28. Sidelobes of bright sources far away from the target field
GB et al. 2010

29. Instrumental spectral response
Pober et al. 2013

30. 3C197.1: ~6.8 Jy
Solutions every 10 sec after
averaging the visibilities over ~230
channels
rms residual: ~9.8 mJy
Calibration accuracy: <0.2%
4C+46.17: ~6.2 Jy
Solutions every 10 sec after
averaging the visibilities over ~230
channels
rms residual: ~6.2 mJy
Calibration accuracy: <0.2%
6C B075752.1+501806: ~5.8 Jy
Solutions every 10 sec after
averaging the visibilities over ~230
channels
rms residual: ~6.2 mJy
Calibration accuracy: ~0.4%
Ionospheric distortions
GB et al. 2010

31. Results

32. Giant Metrewave Radio Telescope (GMRT)
• Large collecting area;
• Clever calibration strategy
(pulsar on–off);
• Stable and known beams;
• Small field of view;
• Severe RFI problems;
Paciga et al. 2011, 2013

33. (Low Frequency ARray) LOFAR
• Largest collecting area;
• Complex elements (levels of
dipole clustering)  element
beams are inherently different
from each;
• Small field of view;
• Active RFI environment
(mitigated by high time and
frequency resolution);

34. Deep imaging on 3C196 and NCP fields

35. Murchison Widefield Array (MWA)
• Centrally condensed core to
maximize power spectrum
sensitivity for the EoR (but
smaller collecting area);
• Large field of view;
• Minimum RFI contamination;
• Analog signal paths;
courtesy A. Offringa

36. Upper limits on the EoR at z~8.5 from the 32T prototype
Dillon, …, GB, et al. 2013
Δk < 0.26 K @ 95% c.l.

37. Precision Array to Probe the Epoch of Reionization (PAPER)
• Maximum redundant configuration
(baselines length are equal to each
other as much as
possible), optimized for EoR
power spectrum measurement;
• The simplest design (beam
stability, smoothness, minimal
ionospheric impact);
• Very large field of view;
• Minimum RFI contamination;
• Analog signal paths;
• Smallest collecting area;

38. Upper limits on the EoR from PAPER 32
WSRT (GB et al., 2010)
Courtesy J. Pober

39. What have we learned about reionization from current 21cm
measurements?
Xi ~ 0.5
The IGM must have been heated by X-rays (MXRBs a/o quasars)

40. Hydrogen Epoch of Reionization Array:
HERA-576
http://reionization.org

41. Conclusions
• The redshifted 21cm line promises to be a fantastic probe of the high-z Universe;
• Steady progress towards the first detection of HI at z > 6;
• Many challenges still to be overcome (calibration, foreground subtraction) –
development required!;
• The detection will open up the field for a characterization of the EoR and Dark Ages;
• SKA low and HERA looking ahead;
• Observations of the global sky signal represent a way to probe the cosmic dawn at z
~ 25-30;
• HI intensity mapping!  BAOs at 0.1 < z < 6