The UK ACoRNE Group Present Projects and Future Plans

1
The UK ACoRNE Group Present Projects and Future
Plans
J. Allen, R. Binns, S. Danaher, J. Perkin, C.J.
Rhodes, T. Sloan, L.F Thompson, D. Waters P.
Doust (ISVR Southampton)
Acoustic Cosmic Ray Neutrino Experiment

• Sean Danaher Northumbria University
• sean.danaher_at_unn.ac.uk

2
Summary

• UK Group Funded
• MC simulation studies
• Energy Deposition Simulator
• Acoustic Parametric simulator
• Upgrade of Rona
• Future Plans
• Conclusion

3
UK Group

• Collaboration
• Funded for 3 years Jointly by PPARC and the MoD
• OKd Sept 04
• In place Feb 05
• RA Job advertised

4
Rationale
Monte Carlo Studies Corsika Geant IV

• Full Amplitude and Phase calibration of Array
• Pulse Simulator
• Parametric Acoustic Simulator
• Amplitude and Phase Responses of system

Acoustic Integrals Improve speed Include
fluctuations Real attenuation
Data Gathering and analysis Gather a large
dataset of unfiltered data Use numerous
techniques e.g. Signal processing classification
algorithms to separate signal from noise and
background
Butterworth Filter pulse Response
Understand scaling to large arrays Be able to
make recommendation for a full size experiment
within 3 years
Saund light bulb
time (ms)
5
Shower Simulation Studies

• Using GEANT IV and CORSIKA to model the shower
  shape both Longitudinal and average Radial
• Wish to extend to E(r,l) Radial distribution as a
  function of shower age
• Use E(r,l) to design an accurate simulator

6
Acoustic Detection Principles

• Energy must be deposited over the same volume as
  the shower
• On axis pulse approximates derivative of cross
  section of shower (a few cm)
• Angular Spread depends on wavelength the length
  of the shower (?/d0.15/10 1O)
• Energy must be deposited within c 10?s
• (100kHz)

7
Light Deposition Simulator
Gaussian Lateral distribution, Exponential
longitudinal decay
Absorption Coefficient of pure sea water
0.7
0.6
gt1J in Green expensive! Use optics to spread
beam. Access to MoD laser (5J at 527nm)
0.5
0.4
High Power Leds 100ns rise time 5W continuous in
Green Pulsing Circuits
Absorption Coefficient(m-1)
0.3
0.2
0.1
Xenon Flash Guns 1000J available White. Use
focussing optics to mimic spread of Shower.
0
400
450
500
550
600
650
700
Wavelength (nm)
8
Acoustic Parametric Simulator
Receiver equalisation
Wideband power amplifier
Output equalisation (Analogue)
Surface anechoic cladding.

• Can put bipolar pulses in water
• Can retrieve bipolar shape
• Rona array to be categorised over the summer
• Investigating linear hydrophones

1m
D140 Hydrophone
2m
D17 TX
5m
Storage Oscilloscope
Electrical input signal.
Input equalisation (Analogue)
Signal generator
HF monitor hydrophone.
Electrical input signal.
Unequalised transmitted acoustic waveform.
LF monitor hydrophone.
HF monitor hydrophone.
LF monitor hydrophone with equalisation.
Equalised transmitted acoustic waveform.
P Doust ISVR Southampton
9
Acoustic Parametric Simulator II
The Group has acquired a line array this can be
used to beam form the sound into the
Pancake String Hydrophones also under
investigation
Can Also be used for beam-forming (in receive
mode) Factor of at least 8dB
10
Sheffield Tank

• Currently commissioning
• Hydrophones
• Data Acquisition system
• NI DAQ Card-6062E (for PCMCIA)
• 500 kS/s, 12-Bit, 16 Analogue Input
• BK 8104 Hydrophone
• (?4.0 dB) 0.1 Hz to 80 kHzwith BK Amplifier

11
Rona Hydrophone Array

• MoD facility in North West Scotland
• An array of high sensitivity hydrophones with a
  frequency response appropriate to acoustic
  detection studies
• Existing large-scale infrastructure including
  DAQ, data transmission, buildings, anchorage
• Phase to be calibrated in near future
• PPARC/MoD funding permits us to upgrade Data
  Acquisition system there to permit several weeks
  worth of unfiltered data to be recorded
• Provides an excellent test-bed for the simulator

12
Knusden Noise and Matched Filtering
Is one quiet day better than many noisy days?

• Noise driven primarily by sea state
• Can Model Noise Spectrum as H(z)
• Design optimal Matched Filter

For a given site can model sea state noise using
a Rayleigh Distribution
13
Matched Filtering

1. The signal sits on a flat noise background
   varying only in magnitude according to sea state
2. Apply matched filter 3 times noise reduction
3. Limit PFA to 1 signal in 10 yrs due to noise with
   5 fold coincidence gives us a threshold with
   POD 50 of typically 35 mPa

14
Advanced Signal Processing Techniques

• Wavelets
• Adaptive Fuzzy Clustering
• Subspace Optimisation (SVD)
• Artificial Neural Networks
• MLP
• Kohonen
• LVQ
• Adaptive filtering (Wiener etc)

15
The Future?

• Rona very useful as a test bed
• But ideally
• Hot, Salty, Deep, low average wind velocity low
  shipping noise E Med, Mid Atlantic?
• Need multi-parameter detector with data fusion?
• Need an EU/US/Russian Collaboration?

16
ARENA 2006Newcastle UK