detector modelling in astroparticle physics -...

6th International Workshop Data Analysis in Astronomy«Livio Scarsi»

“Modelling and Simulation in Science”Ettore Majorana Foundation and Centre for Scientific Culture

Detector Modelling in Astroparticle PhysicsDetector Modelling in Astroparticle Physics

Sergio Petrera, INFN and L'Aquila University

Particle vs AstroPart. PhysicsDetector ModellingTwo examples:

Atmospheric neutrinos (MACRO)UHE Cosmic Rays (Auger)

Particle vs Astroparticle Physics (my personal view)Particle vs Astroparticle Physics (my personal view)

Both exploit particle detection...but, the physics aim is different

In Particle Physics the beam beam (beam-target, beam-beam) is perfectly knownThe interest is addressed to the outcome of the interaction: e.g.

e+ e-

Z0

Particle vs Astroparticle Physics (my personal view)Particle vs Astroparticle Physics (my personal view)

Both exploit particle detection...but, the physics aim is different

In Particle Physics the beam beam (beam-target, beam-beam) is perfectly knownThe interest is addressed to the outcome of the interaction: e.g.

In Astroparticle Physics the beam is unknown beam is unknown (partly or fully)The interest is addressed to the knowledge of its features: e.g.

particle identityenergydirectionorigin

γ

... their common feature: particle detectionparticle detection

Detector modelling usually done with dedicated codes:Experiment specific simulation programGeneral purpose packages/toolkit (GEANT)

Simulation includes:Detailed geometry of detectorsGeometry of passive elementsPhysics processes originating signals...

Let's look at GEANT

Detector Modelling in AstroParticleDetector Modelling in AstroParticle

There are specific features:



Apart astroparticle physics searches in space, the observation is always indirectalways indirect.



Apart astroparticle physics searches in space, the observation is always indirectalways indirect. This means that one has to infer primary physics parameters from

secondary particlessecondary particles.




secondary particlessecondary particles. Detector modelling is then extended to the surrounding environmenextended to the surrounding environment

where the secondaries are developed




secondary particlessecondary particles. Detector modelling is then extended to the surrounding environmenextended to the surrounding environment

where the secondaries are developed

Examples of this kind are e.g.neutrino detectors in underground experiments (+ rock)ground based cosmic ray experiments. (+ atmosphere)

Case I: MACRO Case I: MACRO

The beam: atmospheric neutrinos



The physics phenomenon: νμ→ντ oscillation




How do we see it? νμ as a function of angle ϑ





How do we detect νμ?






The role of detector simulation is fundamental!Response of the detector (tracking, timing, etc.)Neutrino induced muons (flux, cross section): the rock surrounding

the detector becomes part of the simulationMuon transportBackground evaluation: fake upward muons (...again the rock...)






The role of detector simulation is fundamental!Response of the detector (tracking, timing, etc.)Neutrino induced muons (flux, cross section): the rock surrounding

the detector becomes part of the simulationMuon transportBackground evaluation: fake upward muons (...again the rock...)

The detection and measurement of neutrino induced muons was very robust in MACRO. The oscillation search would have been impossible without anThe oscillation search would have been impossible without anaccurate simulation!accurate simulation!

Simulating a Large Cosmic Ray Experiment:

The Pierre Auger Observatory

T. PaulNortheastern UniversityBoston, USA

An example of some ideas described at this workshop for case of one experiment

Pierre Auger ObservatoryObjective: study the highest energy cosmic rays

What is the shape of the energy spectrum ?What is the composition Where do they come from(point sources, anisotropy) ?

Observatory:Large exposure to capture rare eventsEach hemisphere for full sky coverageComplementary detection methods

km 2 sr

Mendoza, ArgentinaNearing completion

Colorado, USAIn planning stages

aperture ~7000

Nitrogen fluorescence detectedas shower develops

Particles detected as they reach ground

Detection Techniques

Fluorescencenearly calorimetric direct view of shower evolution10% duty cycleAcceptance depends on energy + atmosphere

Surface100% duty cycleFlat acceptance above thresholdIndirect measurements of primary energy and mass (relies on simulation)

Hybrid = surface + fluorescence

Southern observatory

1600 detectorscovering 3000 km2

Surface array

Fluorescence detectors4 buildings6 telescopes each

Surface detector

12 tons purified waterradiates Cherenkov light

3 Photomultipliertubes

Tyvek bag actsas diffusive reflector

Communicationantenna

GPS antennafor timing

Electronicsenclosure

1.2m

1.8m

Fluorescence telescope

440 pixel camera

aperture and filtermirror

mirror

Fluorescence buildings (“Eyes”)

Atmospheric monitoring

Laser facilities

LIDAR at each eye

Understanding fluorescence data requires monitoring atmospheric conditions

Weather stations& Radiosondes

The role of simulationThe role of simulation

Is crucial for:CR interaction and shower development Photon transport to FD telescopesDetector simulation (in SD tanks, ray tracing to PMT pixels)

Even in this case detector simulation extends to the surrounding extends to the surrounding environment:environment: the atmospherethe atmosphere

A graded approach for each basic detector:Fast simulators with home made codeGeant4 fast simulation Geant4 full simulation

The most time consuming simulation used as a reference

Detector simulations used for :Inferring composition from measured quantities(see talk from J. Knapp)

Shower front curvature(from arrival times at each station)

depth of shower maximum

photons

protonsiron

depth of shower maximum

riset

ime

curv

atur

e (k

m)

Risetime and curvature determined from surface detector measurements!Simulations are required to relate these measurements to properties of the primary.

Shower risetime(from signal timing in surface detectors)

ironprotons

photons

photons

Detector simulation use for (cont.)Developing and validating reconstruction algorithms(reconstruction = turning measured quantities into shower properties)

Estimating systematic errors

Computing the fluorescence detector exposure

Design studies for enhancements (or whole new observatories)

efficiency: depends on energy, atmosphere, detector issues ...

changes duringconstruction

Crosschecking calibration procedures

Steps for simulating surface arrayEach simulation step encapsulated in a module

corsikaairesconexseneca

shower core placement

Shower Unthinning

phot

oele

ctro

ns

time

Tank response to particles

time bin

coun

tsPhototube and electronics

Station triggeringvarious combinations:coincidencesthresholdtime-over-threshold

Central triggeringspace-time clusteringof tank signals

Event building40 MHz sampling

(particles hitting tanks)

Example: simulating tank response

Particle interactions in tank materialCherenkov light generated in water

enteringentering

ray productionenergy loss

Tyvek reflectivity set to 0

Charged particles emit Cherenkov light

Cherenkov light bounces off Tyvek

Entering particles

Tank properties tuned to measurementsReflectivity vs.Specular / diffuse

refection

Absorption in water vs. Optical interfaces at phototubePhototube quantum efficiency vs

Optical photons

Tank model vs. dataTank models are verified with data.

inclined muons

Example: muon hodoscope used to trigger on atmospheric muons

“vertical” muons

Details like this are important for understanding detector response vs. zenith

scintillatorpaddles

Tank model vs. dataTanks self-triggering by phototube coincidence

signal sizesignal size

Signal shape affected by:Different trajectories through tanks make different signal sizesEnergy and angular distributions

Steps for simulating fluorescence telescopescorsikaairesconexseneca

FluorescenceCherenkov

Shower light sim Light propagation up to diaphragmPropagation:

Direct fluorescenceattenuation

Indirect Cherenkov:Rayleigh scatteredAerosol scattered

Raytracing in telescope

time bin

coun

ts

Electronics simulation

photons to ADC countsnoise

Trigger simulation

Event Building

Calibration sim

Background sim

dEdx

Geant4 modellingof the FD telescope

A visualization in Geant4:

ImplementationSome Details Mercedes light collectors

Assembled camera

PMTs described as hexagonal volumes (shown in red) and defined as sensitive volumes.

Simulation and reconstruction(surface and fluorescence)

shower direction

longitudinal profile

depth of maximum (related to composition)

lateral profileshower front shapesignal risetimes

simulation data acquisition format reconstruction

SummarySummary

Astroparticle physics experiments derive detector Astroparticle physics experiments derive detector simulation from the wide experience of HEP simulation from the wide experience of HEP accelerator experimentsaccelerator experiments

Generally detectors are simpler and less numerous Generally detectors are simpler and less numerous than in modern accelerator experimentsthan in modern accelerator experiments

Nonetheless the detector simulation needs the Nonetheless the detector simulation needs the insertion of the detector surrounding material to insertion of the detector surrounding material to reproduce the datareproduce the data

As for physics generators big progresses in the As for physics generators big progresses in the last 10 yearslast 10 years

The end

detector modelling in astroparticle physics -...

Documents