Simulation Issues for Radio Detection in Ice and Salt

1
Simulation Issues for Radio Detection in Ice and
Salt

• Amy Connolly
• UCLA
• May 18th, 2005

2
Overview

• During the past few years, simulations of
  Askaryan pulses and detection systems have become
  mature
• Large overlap in people working on simulations
  from GLUE, ANITA, SalSA
• This talk focuses on the status of simulations
  developed for ANITA, SalSA
• Lessons learned

3
Simulations - Overview

• Complementary simulation programs are being
  developed this is essential
• Interactions occur uniformly and isotropically in
  medium followed by weighting

• Flavors treated separately
• EM, Had components of shower treated separately
  for simulating Askaryan signal, proper dN/dy used
• Secondary interactions included
• Weighting accounts for neutrino attenuation
  through Earth, etc.

P. Gorham
4
Anita Simulation

• Ray tracing through ice, firn
• Fresnel coefficients
• Attenuation lengths are depth and frequency
  dependent
• Include surface slope and roughness
• 40 quad ridged horn antennas arranged in 3
  layers 8,8,16
• Bandwidth 200 MHz-1200 MHz
• For now, signal in frequency domain
• Measured antenna response

S. Barwick
5
Anita Ice Properties
Use measured ice and crust layer thicknesses
(model is Crust 2.0 based on seismic data)
n(z) 1.8 in deep ice 1.3 at surface
D. Besson
Rays traced in ice with depth-dependent index of
refraction.
P. Gorham
Use measured attenuation lengths (frequency
dependent)
Temperature-dependence also included few
hundred m in warmest ice (in firn and near
bedrock) 1300 m at mid-depth
6
ANITA - Skymap
For a fixed balloon position, sensitivity on sky
takes a sinusoidal shape. A source between -5 and
15 declination would be observable for 5
hrs/day
Over the entire balloon flight, sensitive to
entire band between -10 and 15 declination.
7
Reflected RaysWork by S. Barwick, F. Wu from
University of California at Irvine

• ANITA could (possibly) detect events where a
  signal is reflected from ice-bedrock interface

• At large cross-sections, short pathlengths ?
  down-going neutrinos dominate ! reflected rays
  important

S. Barwick F. Wu
Micro-black holes at
ANITA Energies
S. Barwick

• Signals suffer from extra attenuation through
  ice and losses at reflection
• At SM ? s, reflected rays not significant

1
100
10
1000
?/?SM

• Could measure cross section from relative rates
  of direct (far) to reflected (near).

8
Reflected RaysWork by S. Barwick, F. Wu from
University of California at Irvine
With reflected rays, we could observe a large
portion of the sky that we could not otherwise.

• HOWEVER, more uncertainty at ice-bedrock
  interface
• For now, assuming 10 attenuation in power at
  interface
• Collaborating with UT group to understand
  under-ice topologies, radar reflectivities
• Use Brealt code to study interfaces
  quantitatively

9
SalSA Benchmark Detector Parameters

• Overburden 500 m
• Detector
• Array starts at 750 m below surface
• 10 x 10 string square array, 250 m horiz.
  separation
• 2000 m deep, 12 nodes/string, 182 m vert.
  separation
• 12 antennas/node
• Salt extends many atten. lengths from detector
  walls
• Attenuation length 250 m
• Alternating vert., horiz. polarization antennas
• Bandwidth 100-300 MHz
• Trigger requires 5/12 antennas on node, 5 nodes
  to fire Vsignalgt2.8 ? VRMS
• Index of refraction2.45
• Syst. Temp450K300K (salt) 150K (receiver)

10
Angular ResolutionWork by P. Gorham,
University of Hawaii

• Performed chi-squared analysis from two hadronic
  shower event types
• Fully contained
• Parallel to a face 250 m outside array
• Fit to
• Amplitude of Cerenkov signal
• Polarization
• At 81016 eV
• Contained fraction of deg.
• Non-contained 1 deg.
• Improves with energy

P. Gorham
11
SalSA Cross Section Measurement
? N
cross section can be measured from
distribution
cos ??
Generate distribution from
simulation and throw dice for many
pseudo-experiments
Binning 2
_at_1018 eV,
For experiments w/ 100 events
ltLmeasintgt400 130 km
With 300 events
At SM ? , only 10 of events in sensitive region
Use Poisson likelihood
12
Comparing Sensitivities

• SalSA ANITA
• SalSA lower threshold
• ANITA higher V?, shorter livetime
• Two independent simulations for each experiment
  give similar results for ESS baseline
• ANITA handful of events
• SalSA 20 events/year
• Having two MCs has been essential to
• Learn about the physics of these systems
• Spot bugs
• Gain confidence in our results

RICE 333 hrs
GLUE 120 hrs
SalSA (1 year)
ANITA 45 days
ESS baseline
13
How Do We Know Our Simulations Are Correct?
Even if two independent simulations give the same
answer, we should assume it is a coincidence
until we compare many, many intermediate plots.
ANITA
Resolving disagreement in peak at 11020 eV
Depth of Interaction (m)
Balloon-to-Interaction Distance (km)
14
Validating our Simulations (cont)
ANITA
Projection of Askaryan signal onto the sky
S. Hoover
P. Gorham
15
Summary

• Simulations of radio detection systems are
  becoming sophisticated
• ANITA
• Reflected rays show promise for detecting high
  cross sections, opening up large part of the sky
  if ice-bedrock interface can be understood
• SALSA
• Angular resolution fraction of degree for
  contained events, 1-2 degrees for external events
• Cross section measured at 30 level with 100
  events
• Independent simulations are essential
• Many intermediate plots necessary to verify
  simulation performance

16
Backup Slides
17
ANITA
SalSA
GLUE
RICE
18
Impact of Salt Properties

• Track length L
• X0ice43 cm, X0salt10.3 cm !
  Expect Lsalt/Lice0.26. Simulations show 0.34.
• Cerenkov index of refraction factor
• Cerenkov threshold
• Critical energy
• Coherence
• Angular scaling

19
Secondary Interactions

• Generate from MMC for each
  flavor, interaction type
• From MMC, also retrieve multiplicity of each type
  of sec. interaction
• Force neutrinos in our simulation to obey these
  distributions
• For now, consider interaction (primary or
  secondary) which contributes the strongest signal
  
• Critical for flavor ID

Example Probability Distribution from MMC for
muon brem. showers
P. Miocinovic