1. Lunar Radio Cerenkov Observations of UHE Neutrinos
Ron Ekers
CSIRO
ARENA 2008
Rome, June 2008
Invited talk at XXIII Texas Symposium on Relativistic Astrophysics 2006
23 June 2008

2. Lunar Cherenkov emission
neutrino
photon
shower

3. Summary History of lunar Cerenkov experiments Basic technology
Low or High frequency
Telescope arrays
Future developments and UHE neutrino sensitivity
23 June 2008

4. First radio experiment Parkes 64m radio telescope
Jan 1995
Triggered by Adelaide ICRC meeting
Receiver:
1.2 – 1.9 GHz. (SETI receiver)
Used existing pulsar equipment
Bandwidth:
475 MHz
2 sec time resolution.
Beamwidth:
13 arc min.
Moon ~ 30 arc min, hence reduced sensitivity at Moon’s limb !
Hankins, Ekers & O’Sullivan MNRAS 283, 1996

5. Basic requirements Radio receivers (0.1-3GHz) Sensitivity
10-300cm
Sensitivity
aperture >> 20m diameter
Observing time >> weeks
GHz bandwidth and η sec time resolution
Radio Frequency Interference (RFI) discrimination
Good site
Real time pulse detection logic and data recorder
Post trigger data processing
23 June 2008

6. Terrestrial Interference
Forte satellite: 131MHz
FORTÉ satellite: MHz
10 Dec 2006

7. Pulse dispersion effects
 t =  .  ne dl . -3
ionosphere
t = 20 sec
interstellar medium
t = 10 sec E M
Galactic radio pulse t = 10 sec
Moon bounce t = 2x 20 sec
Lunar Cherenkov t = 20 sec
Terrestrial interference t = 0 -(
10 Dec 2006

8. Ionospheric Dispersion Estimation
TEC (total electron content) is derived from dual frequency GPS
Provided on line by NASA crustal dynamics program
Not available in real time
Some model error for a given line of site
Use simultaneous measurement of the Faraday Rotation of polarised lunar limb emission
Model field distribution => TEC in realtime
23 June 2008

9. Goldstone Lunar Neutrino Search 1998
GLUE
NASA 70m DSN antenna
Peter Gorham, David Saltzberg, Dawn Williams
First observations late 1998:
approach based on Hankins et al results from Parkes
1999: add 2nd 34 m antenna DSS13
Dual trigger from both antenna
Removes local interference
updated
10 Dec 2006

10. Neutrino Flux Limits Goldstone
~30 hrs Goldstone
No events above net 5 sigma
Gorham, Saltzberg et al RADHEP-2000, 177 (2001).
New Monte Carlo estimates:
Full refraction raytrace, including surface roughness, absorption
Parkes Sensitivity
2.5x Goldstone
But only 1hr on limb
2003: 120 hours on source
Gorham et al, Phys Rev Lett 93, (2004)
Updated version
needs more updates with new (GLUE) sensitiovity limits which have changed and which are still wrong
~30 hrs Goldstone
No events above net 5 sigma
10 Dec 2006

11. Historical Limits on total UHE neutrino flux
James and Protheroe arXiv: v2
23 June 2008

12. Detecting the pulse Minimum neutrino energy
Determined by pulse detection threshold, hence radio telescope sensitivity
GHz frequencies better
Moves curve to left
Neutrino flux limit determined by
Volume of neutrino detector
Acceptance solid angle
Lower frequency better
Observing time
Moves curve down
10 Dec 2006

13. pulse detection threshold
Signal
antenna gain proportional to antenna area but....
bandwidth
Noise
thermal emission from moon
Moon unresolved (7m-22m) – increases with antenna size
Moon resolved (>20m) - independent of dish size
Correlated between elements
sky noise
becomes dominant at about 100MHz (moon then colder than sky)!
receiver noise
not very important for this experiment
Ionospheric dispersion
can be fully corrected and can help if few bit sampling used
will increase detection threshold if too uncertain
23 June 2008

14. Sensitivity v Antenna Diameter
Modify title
10 Dec 2006

15. Balancing trigger and post processing sensitivity
At GHz bandwidth raw data rates exceed current computer capacity
x10 now but decreasing
A trigger and data dump strategy is necessary
The experimental sensitivity is almost entirely determined by the trigger level
Requires real-time coherent ionospheric dispersion
Requires sophisticated high speed real-time logic
23 June 2008

16. Some Hardware Issues Wideband transistor receivers
Usually cryogenically cooled in radio telescopes
Dual orthogonal polarizations (linear or circualar)
Wide bandwidth system
Avoid reflections (unwanted modes)
Fast detection logic
State of art FPGA’s
Careful circuit design
23 June 2008

17. Sensitivity v Antenna Diameter
multiplex
Modify title
10 Dec 2006

18. Seeing all the moon Event rates highest at the limb
30’
Event rates highest at the limb
Noise contribution from all moon
We want to observe the entire lunar limb
20m dish
64m dish
array
23 June 2008

19. Imaging antenna

20. Beamforming Array  t Phased array ( Vi )2 I() ( Vi )2 I x2
Split signal
no S/N loss
t
Phased array
( Vi )2
I()
( Vi )2 I

21. Coherent Beamforming Arrays
Noise dominated by thermal emission from moon
Increasing antenna size doesn’t help
Aperture plane or Focal plane
Arrays of small antenna (eg ATCA-SKA)
Form multiple beams from multiple antennas
Coherent addition of signals
Enough beams to cover all moon
Multiple beams in a large dish
Eg Parkes + focal plane array
Max signal from limb of moon
Needs fully sample focal plane array
Beam former to illuminate whole limb
New version – with corrected beamforming
23 June 2008

22. Radio Experiments some issues to consider
Broadband continuum pulse
1-3GHz , Δt = 0.5 ηsec
Ionospheric dispersion
10-50 ηsec at 2 GHz, μsec at 0.2 GHz
Coherent de-dispersion necessary
Avoid solar maximum?
Linearly polarised signal
Measure both orthogonal polarizations
φ gives position around cone, hence position on sky
Pulse detection on each beam at 0.5nS resolution
Real time dump and store events
Coincidence detection for Interference rejection
Pulse coincidence detection for spaced antennas
Anti-coincidence using multiple beams in one antenna
modified
23 June 2008

23. Position determination
Interferometer measures position of pulse
Source lies on the Cherenkov cone (56o)
Linear polarization with radial distribution *
Includes intensity variation around the cone which removes the amiguity
Measure polarization
Two possible positions
One less probable
10 Dec 2006

24. Radio Lunar Cerenkov Experiments
Parkes 64m 1.5 GHz 1995
Goldstone 70+32m 2.2 GHz 1998-
Kalyazin 64m GHz 2005
WSRT 14x25m 0.15 GHz 2006 observing
ATCA 6x22m 1.5 GHz 2007 observing
LOFAR phased array 0.1 GHz Future
ASKAP 30x12m 1.0 GHz Future
SKA 1500x20m 1.0 GHz Proposed
ELVIS lunar orbiter Proposed
LORD lunar orbiter Proposed

25. Effective Aperture
James and Protheroe , 2008
23 June 2008

26. Low Frequency Experiments <500MHz
Scholten et al astoph/
Weaker pulses but large acceptance angle
Higher energy cutoff eV
Flux limit lower (x10-100)
Many issues
Modest bandwidth so easier signal processing
Cheaper collecting area
Strong assumption about depth of regolith (500 v 10m)
UHE Cosmic ray “contamination”
No positional information
Huge ionospheric dispersion correction needed
Radio frequency interference

27. Square Kilometre Array 2015-2020
23 June 2008

28. SKA Aperture Array 70-500MHz
23 June 2008

29. SKA Poster
10 Dec 2006

30. SKA configuration Trigger off core Keep long baselines in a buffer
Inner core
Station
Keep long baselines in a buffer
Combine high and low frequency triggers

31. UHE neutrino models and future experimental limits
James and Protheroe arXiv: v2
23 June 2008

32. China FAST Karst region for array of large Arecibo-like Telescopes
Diameter 600 m
Add a video clip

33. Phased Array Feeds
Full aperture
Whole moon
23 June 2008

34. Installing the Parkes 21cm Multibeam Receiver
10 Dec 2006

35. Conclusions
Fruitful interaction between radio astronomy and high energy particle communities
Some culture differences
Big improvements in signal processing technology
Significant sensitivity improvements since 1995
Sensitivity estimates have been improved
Sensitivity estimates don’t all agree (many assumptions)
High v low frequency?
Future sensitivity improvements will need antenna arrays or focal plane arrays
Proposed future radio facilities will probe all UHE models
Strategy for targeted experiments is different
23 June 2008