ACoRNE Status Report

1
Status report
2007
2
Introduction ACoRNE collaboration institutions

• The Acoustic Cosmic Ray Neutrino Experiment is a
  UK collaboration represented by researchers at
• Imperial College London
• The University of Lancaster
• The University of Northumbria
• The University of Sheffield (Coordinator - Lee
  Thompson)
• University College London
• The project is funded through JGS by PPARC and
  DSTL and allows us access to QinetiQ acoustic
  range off Rona in NW Scotland.

3
Contents ACoRNE key skills / deliverables

• Data Acquisition
• Rona acoustic range
• ACoRNE DAQ philosophy
• ACoRNE filter chain
• Hardware Development
• Calibration devices
• Signal Processing and Data Analysis
• Rona DAQ
• Offline filtering and analysis
• MC simulations
• Hadronic shower simulations
• Large scale future detectors and hybrid arrays

4
The Rona underwater acoustic range

• Comprises 8 wideband hydrophones ITC-8201 flat
  -158 dB/v/µPa from 0.01-65kHz, omnidirectional
• Distributed around a rectangle 1200m x 200m at
  mid depth in 230m deep water
• Readout is 16Bit _at_ 140kHz
• Rona has a facility to locate hydrophone
  positions via acoustic beacons but this is
  offline when we take data

MOVIE
5
ACoRNE DAQ philosophy

• All data to shore - unfiltered
• All processing and analysis offline
• 8TB RAID interfacing to LT03 tape robot
  (400GB/tape) with 16 tape autoloader and barcode
  reader
• This DAQ system is under the ownership of the
  ACoRNE collaboration and in essence can move to
  another location should the opportunity arise
• FLAC lossless codec providing 55 compression on
  WAV audio files
• Linear phase filtering enabled
• Unlimited ability to reprocess raw data
• Filters evolving over time

6
ACoRNE filter chain

• Currently, raw data is passed through reduction
  process encompassing a series of filters designed
  to optimise retention of neutrino-like signals
• 4 types of trigger in order of increasing
  priority
• T1 P
• T2 dP/dT
• T3 d2P/dT2
• T4 M, where M is one or more matched filters
• Data analysed in 10s intervals
• Top 5 of each trigger per interval saved
• 1000 samples either side of trigger time stamp
  type

7
ACoRNE filter chain

• Now on second generation of filters
• 1st generation used single matched filter which
  was optimised for bipolar acoustic pulses at 1km
  from range, in the plane of the acoustic pancake
• 2nd generation uses 9 matched filters optimised
  for varying distance up to 2.5km from centre of
  range and 2.5 out of the pancake plane
• Given the dimensions and angular acceptance of
  the array to pancake like emission can optimise
  for distances and angles detectable neutrinos are
  most likely to inhabit

8
Rona data

• First campaign 9th-24th December 2005 2.8TB raw
  data (_at_55 compression) 1 matched filter in
  reduction
• 12.453x106 events retained after passing
  through reduction
• Novel analysis performed, requiring gt 4fold
  coincidence above 35mPa threshold
• Events categorised in terms of Fourier spectra
• Second campaign in progress, currently amassed
  15TB (_at_55 compression)
• Analysis underway, following further development
  of matched filter

9
Hardware development calibration devices

• Meaningful astronomy via the acoustic detection
  of UHE neutrinos relies on our ability to
  calibrate hydrophones
• The perfect calibration tool would deposit
  thermal energy in the sea with equivalent
  dimension and density to a neutrino induced
  hadronic cascade

• Hence generating acoustic pancake from
  thermoacoustic emission
• High power (gt1J per lt100µs pulse) lasers could
  facilitate this requirement, availability and
  deployment are limiting factors

10
Hardware development calibration devices

• An equivalent signal to that produced by the
  thermacoustic mechanism can be produced via
  acoustic transducers
• First stage in this process is to generate a
  bipolar acoustic pulse from a single,
  omnidirectional element
• Next step involves a linear array of transducers
  emitting in phase to create an interference
  pattern analogous to the acoustic pancake
  resulting from a hadronic cascade in the sea
• More on this in calibration talk.

11
Hardware development calibration devices

• This summer the ACoRNE collaboration will deploy
  an omnidirectional, single element calibration
  source above the Rona acoustic range
• For the first time we will be able to test the
  emission and detection of a neutrino-like bipolar
  acoustic pulse in situ and in the presence of
  noise
• Thus far tests have been performed in the ACoRNE
  tank, the Sheffield University swimming pool,
  and, Kelk lake (East Yorkshire)

12
Signal processing and data analysis

• A novel analysis of the first 2.8TB of Rona data
  from the first campaign was undertaken (J.Perkin
  thesis)
• Simple coincidence and threshold cuts, in
  accordance with large scale detector simulation
  (discussed later) and analysis of Fourier spectra
  were used to classify different types of
  recurring event in the data
• Threshold 35mPa
• Coincidence gt 4 hydrophones
• Coincident signals were required to show same
  features in DFT spectrum
• 3526 of the 12.453x106 reduced events met passed
  cuts

13
Some classes of event observed at Rona
50kHz oscillating signal
140kHz sample rate N27128 sample FFT
5kHz oscillating signal
Pressure (mPa)
Short-lived impulsive event
Sample No.
14
Some classes of event observed at Rona
140kHz sample rate N27128 sample FFT
10kHz oscillating signal
A ringing event
Pressure (mPa)
Sample No.
A bipolar event
15
Offline filtering and analysis Matched triggers

• The matched trigger will try to pattern match the
  pulse shape
• It is optimised to Knudsen Noise
• Plotted here is the derivative of 1000 triggers
  and the derivative of our expected bipolar pulse
• Note time delay of filter causes the pulse
  maximum to be at -70µs

16
Offline filtering and analysis variation of
triggered events with time
No obvious correlation with time
Normalised event frequency
Day in Dec 2005
17
Offline Analysis

• 13 dimensional phase space to explore
• Pulse Width
• Pulse Periodicity
• Relative Energy
• Pulse Multiplicity
• Dominant Frequency
• Sinusoidalness
• Bipolarity
• Standard Deviation
• Skewness
• Kurtosis (Gaussianisity)
• Assymetry of Standard Deviation
• Assymetry of Skewness
• Assymetry of Kurtosis
• Prioritisation/optimisation of phase space
  parameters is underway.

18
Monte Carlo Simulations hadronic cascades in
seawater (and ice)

• Paper submitted to Astroparticle Physics
  arXiv0704.1025
• Developed an extension to the CORSIKA air shower
  program enabling the simulation of UHE neutrino
  induced hadronic cascades in water and/or ice
• Validated via comparison with Geant4 (Elt100TeV)
  and then compared with existing parameterisations
  (NKG, Niess Bertin)
• Parameterisation of modified CORSIKA results
  proposed to allow for fast computation of thermal
  energy density - for use in large-scale detector
  simulations

19
Modified CORSIKA verifying LPM effect

• Firstly varify LPM effect is implemented in
  CORSIKA
• Suppresses PP and Brems in EM showers
• Only energy density of hadronic part of
  interaction sufficient for thermoacoustic emission

20
CORSIKA vs Geant4 at 1-100TeV

• Next verify that modified version of CORSIKA
  agrees with Geant4 to validate modifications made
  to CORSIKA
• Agreement within 20-30 accuracy of calculated
  neutrino-nucleon cross sections / physics models

21
Modified CORSIKA from proton induced showers to
neutrino induced showers

• The thermal energy density resultant from a
  single high energy proton is the same as that
  resulting from a UHE neutrino interaction
  multiplied by a scale factor

22
The acoustic pulse from a CORSIKA generated shower

• The acoustic signal 1km from shower axis in the
  pancake plane
• Average of 100 CORSIKA showers at 109 GeV in
  water
• The curves show the deposited energies within
  cores of radius 1.025g cm2, 2.05g cm2 and the
  whole shower respectively
• It can be seen that most of the amplitude comes
  from the energy within a core of radius 2.05g cm2

23
Comparison of CORSIKA with existing
parameterisations
Good agreement with Niess Bertin (within model
uncertainties)
24
Conclusions from CORSIKA study

• The thermoacoustic energy deposited by UHE
  neutrino induced hadronic cascades can be
  simulated in seawater and ice without
  extrapolating results from lower energy physics
  toolkits such as Geant4
• The acoustic signal generated by a hadronic
  cascade is very sensitive to the core of the
  shower
• Hadronic cascades initiated by protons and
  neutrinos are broadly equivalent if one scales
  for the energy transfer from the neutrino to the
  hadronic component of the event
• A parameterisation of the CORSIKA results to be
  used for fast computation of thermoacoustic
  energy densities has been proposed - a
  combination of modified versions of the
  Gaisser-Hillas (longitudinal) and NKG (radial)
  functions are used see paper for definition

25
MC simulations estimating the sensitivity of
large-scale hydrophone arrays to the flux of UHE
neutrinos

• A simulation toolkit based in C/ROOT has been
  constructed to facilitate estimations of the
  performance of large-scale underwater arrays of
  hydrophones
• The program executes the following steps
• An ensemble of downward going neutrinos is
  generated, with an energy spectrum flat in LogE
• A hydrophone array is constructed in a deep sea
  environment from a file containing the
  coordinates of typically 1000 hydrophones
  distributed at random in 1km3
• The program then samples the neutrino spectrum on
  an event by event basis
• Each neutrino is forced to interact in a fiducial
  volume called the can surrounding the hydrophones
• Those hydrophones that do not intersect with the
  acoustic pancake are cut

26
Sensitivity of large scale hydrophone arrays

• The peak pressure as a function of neutrino
  energy and attenuation at each remaining
  hydrophone is calculated for both refracted and
  unrefracted sound rays, assuming each detector
  has an omnidirectional sensitivity
• The pressure amplitude and arrival time for all
  those hydrophones that register a signal above
  threshold is recorded, taking into consideration
  the matched filter performance
• For those events that register hits above
  threshold reconstruction of the interaction
  vertex and neutrino trajectory is attempted
• If the vertex reconstruction algorithm returns an
  unphysical vertex the event is discarded,
  otherwise it classified as a detection
• The sensitivity of the hydrophone array under
  examination is calculated from the number of
  reconstructed events. This is translated into a
  limit on the neutrino flux assuming that the
  hydrophone array detected no events for a given
  period of observation.

27
Sensitivity of large scale hydrophone arrays

• We want to be able to detect GZK neutrinos,
  therefore we have to think big!
• 1500km3, 100hydros/km3 5yrs, 5mPa threshold

28
Simulation of hybrid arrays

• Within the KM3NeT collaboration a Mathematica
  based simulation toolkit has been developed for
  (fast) calculation of optical neutrino telescope
  performance - NESSY
• NESSY is further optimised for calculation of
  detector cost as well as physics capability
• The ACoRNE collaboration aims to further develop
  the NESSY program to include acoustic detection
  and thus enable the simulation of hybrid arrays

29
Summary of ACoRNE activities

• Over 17TB of data are on tape from the Rona
  acoustic range, new set of offline filters
  integrated into analysis
• Omnidirectional bipolar calibration source ready
  to deploy in sea above Rona range, vessel for sea
  operations is chartered, awaiting confirmation
  from QinetiQ
• A new parameterisation based on results from a
  modified version of the CORSIKA program to enable
  fast computation of thermal energy densities
  resulting from neutrino interactions in water and
  ice has been realised
• An existing code is available for the simulation
  of large scale hydrophone arrays
• Work is underway to include the acoustic
  technique in a new Mathematica based program for
  fast simulation of hybrid detectors

30
Future plans for ACoRNE

• Combine noise data from Rona with MET office data
  and perform coincidence study
• Continue analysis and evolution of offline
  filtering to maximise signal/noise
• Deploy omnidirectional bipolar source in sea over
  Rona range and verify detection capability
• Upgrade omnidirectional calibrator to linear
  phased array and put a pancake-like signal in the
  sea (Ensure this technology is good for 2km
  depth and deployment at alternative sites e.g.
  Nemo test site, Capo Passero)
• Complete upgrade to KM3NeT Mathematica code for
  simulation of hybrid detectors