Transcript Simulating Anita and SalSA

Simulation Issues for Radio
Detection in Ice and Salt
Amy Connolly
UCLA
May 18th, 2005
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
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]
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
n
ice
[S. Barwick]
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]
[P. Gorham]
Use measured
attenuation
lengths
(frequency
dependent)
Rays traced
in ice with
depthdependent
index of
refraction.
Temperature-dependence also
included:
~few hundred m in warmest ice
(in firn and near bedrock)
~1300 m at mid-depth
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.
Reflected Rays
Work 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]
n
TIR
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
10
/SM
100
1000
• Could measure cross section
from relative rates of direct (far)
to reflected (near).
Reflected Rays
Work by: S. Barwick, F. Wu from University of California at Irvine
View more sky!
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
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:
Vsignal>2.8  £ VRMS
• Index of refraction=2.45
• Syst. Temp=450K=300K (salt) + 150K (receiver)
Angular Resolution
Work 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 8£1016 eV:
– Contained: fraction of deg.
– Non-contained: ~1 deg.
• Improves with energy
[P. Gorham]
SalSA Cross Section Measurement
n N cross section can be measured
from cos n distribution
Generate
distribution from
simulation and
throw dice for
many pseudoexperiments
Binning:
~2±
@1018 eV,
For experiments w/ 100 events
<Lmeasint>=400 § 130 km
At SM  , only 10% of events in
sensitive region
With 300 events
Use Poisson likelihood
Comparing Sensitivities
• SalSA & ANITA
– SalSA lower threshold
– ANITA higher V, shorter livetime
RICE: 333 hrs
• Two independent simulations
for each experiment give
similar results for ES&S
“baseline”
– ANITA: handful of events
– SalSA: 20 events/year
• Having two MC’s has been
essential to:
– Learn about the physics of these
systems
– Spot bugs
– Gain confidence in our results
GLUE:
120 hrs
SalSA
(1 year)
ES&S “baseline”
ANITA:
45 days
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
1£1020 eV
Depth of Interaction (m)
Balloon-to-Interaction Distance (km)
Validating our Simulations (cont)
Projection of
ANITA
Askaryan signal
onto the sky:
[S. Hoover]
[P. Gorham]
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 icebedrock 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
Backup Slides
ANITA
SalSA
GLUE
RICE
Impact of Salt Properties
• Track length L
– X0ice=43 cm, X0salt=10.3 cm !
Expect
Lsalt/Lice=0.26. Simulations show 0.34.
•
•
•
•
•
Cerenkov index of refraction factor
Cerenkov threshold
Critical energy
Coherence
Angular scaling
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