GleamGRB: A Physical Model for GRB Simulation

1
Gleam/GRB A Physical Model for GRB Simulation

• INFN Pisa
• Nicola Omodei
• Johann Cohen-Tanugi

2
Overview

• GRB Gleam
• Goals, why GRB, why Gleam
• What does GRBSim do ?
• GRB-Gleam integration
• GRB physical model description
• The Fireball Model overview
• Shells evoluton
• Shock dynamics
• Radiative processes
• How GRBSim works and what does it provide
• MC vs Reco Light Curves
• Future Improvement new design
• Conclusion

3
GRB Gleam

• Goals
• To simulate a source based on an plausible
  astrophysical model
• Detector response to signal
• Why GRB ?
• Short signal, possible to simulate photon by
  photon, no integration over a long observational
  time is needed
• We are interested in the GRB physics
• Variability of the signal
• High energy emission from GRB unknown -gt Only a
  model that does assumption can be used !!
• Trigger and alert issues related to GRB
• Why Gleam ?
• A lot of stuffs were already there
• Monte Carlo propagation
• Reconstruction
• Digitization of the hits
• The orbit and the tilting of the satellite are
  computed
• Good environment for study the response of the
  detector and for helping in the development of
  analysis tools.

4
What does GRBSim do ?

• It feeds the simulation of the detector with the
  photons extracted from the flux coming from GRB
• Draw photon from spectrum
• Common problem for every kind of spectrum (static
  and transient)
• Given a spectrum -gt randomly extracts a photon
  using the spectrum as probability function
• Interval to wait before the next iteration
• Is a problem only for transient sources
• The default definition Interval 1/Rate can not
  work for transient sources

5
Status of GRB-Gleam integration

• The LAT team has set up a complete simulation
  chain
• Simulation of the incoming flux
• Background Albedo, Cosmics, Diffuse gamma
• 2 different GRB model approaches
• physical model from shells to photons
• phenomenological model extrapolated from BATSE
• The detector is illuminated in the correct way,
  taking into account the orbit and the tilting of
  the satellite (FluxSvc)
• Simulation of the Detector is up
• Geant4, Digitization, Reconstruction
• Triggers and GRB Alerts
• The onboard trigger are taken into account
• All the algorithms can be tested on the simulated
  data

6
GRB physical model simulation

• Weve started from a plausible astrophysical
  source model (Fireball Model) that describes the
  temporal behavior of a typical Gamma Ray Burst
• Rapid variable signal (ms)
• Non thermal emission (Synchrotron Inverse
  Compton Scattering)

GRB is a package, distributed via CVS !!
7
The Fireball Model overview
The central engine emits shells with different
Lorentz Factor. The shells collide -gt formation
of shocks wave inside the shells material The
shock accelerate the electrons that emits by
synchrotron (presence of MF, equipartition
hypothesis). The high energy emission is provide
by the Compton Scattering. Observed shape of the
spikes Fast Rise Exponential Decay (cooling
time depends on the energy of the observed
photons)
8
The shells
Afterglow phase Interaction with
ISM Deceleration
Acceleration phase The shell are optically thick
Internal Shock phase The shells are optically
thin, They move with constant LF
9
The dynamics of the shock


(mr, gr) (ms, gr) mr, gr Eint
Accelerate the electron N(g) g-p
Magnetic field Ub eB Eint
10
Emission Model Synchrotron
For one electron
For a distribution of electrons
11
Inverse Compton scattering

• Standard HypotesisThe inverse Compton
  Scattering boost the observed photon by a factor
  of ge
• The inverse Compton spectrum has the same shape
  of the Synchrotron spectrum, but shifted by ge

Synchrotron
Synchrotron IC
12
How GRBSim works
13
GRB Simulations what it provides

• GRBSim provide a spectrum that varies with the
  time.
• Changing the parameters of the model (Number of
  shells, geometry of the shells,) we can
  reproduce the wild variety of light curves
  observed.
• The physical description of GRB allows to
  investigate the signature of different physical
  phenomena.
• The photon list is extracted according to the
  fireball model (Internal Shocks)

14
Validation Calibration of the physical model

• We explore the parameter space obtaining
  different burst with different properties.
• We select a burst duration from the T90
  distribution observed by BATSE
• The bimodal distribution is reproduced.

• We compare the fluences in the BATSE catalog
  with the simulated.
• Scale laws are in this way reproduced (HR vs T90,
  Intensity vs T90)

• This can be useful for
• Calibrating the physical parameters of the model
• We are planning to extend this approach to
  different physical models (External Shock
  scenario, Cannonball model,) !!

Data from BATSE catalog
Simulated Burst catalog
15
Light Curves
Short Bursts
Long Bursts
16
MC vs Reco Light Curves
17
GLAST GRB
18
New Design Overview
Generator

• The code has to be redesigned
• Improve clarity
• Better organization of the classes
• Separation between co- moving frame and observed
  frame
• Physics processes as separate classes
• Extension to other model and to other scenarios
• Different shells geometry descriptions
• External shock model
• Fireworks model (Barbiellini, Longo Celotti)

Shell
Shock
Physical Processes
SpectObj
19
New design The engine
Engine
The engine fill the shock vector, different way
to do it

• 1. Evolution Of the Shells
• Gamma Lorentz Factor are chosen randomly
• The Shells evolve with the time
• The shells collide, and a shock is computed

• 2. Creating the shocks
• The shocks are created at a certain time
• No information regarding the previous shells
  evolution

20
New design The Shells
Shells

• Different geometry of the shells
• Spherical fireball
• Jet fireball (collimated)

Relativistic Beaming
The radiation emitted from a source that is
moving with a Lorenz factor G toward the
observer appears beamed in a cone of aperture
1/ G
21
New design The Shocks
Shocks
The different Geometry of the emitting region
(shell) will be considered
The shocks vector contain all the information
needed to compute the emission. Internal energy
available Density of Particle accelerated by the
shock wave Distribution of electron
accelerated Magnetic field in the shocked
material Thickness of the emitting region
Clear separation between shocks and emission
process -gt The shock accelerates electrons, and
creates a magnetic field. -gt The shock doesnt
emit radiation !!
22
New design The Physical processes
They have to be separated from the shock.
General interface for radiative processes ?
Physical processes

• Accessibility to the physics (spectrum) both in
  the co-moving frame (shell) in observer Frame
• Computation of the flux in the shells frame
• Lorentz transformation to the GLAST frame.

Distribution of the accelerated
electrons Magnetic Field Geometry of the emitting
region (shell)
Physical process In the shell Frame
Lorentz Transformation Shell(G??) to GLAST
Physical process GLAST frame
23
New design The SpectObj
Spectrum Object
The Spectrum Object is an interface that carries
all the information and methods that permit the
manipulation of a spectrum Algebra of fluxes
( - /), units conversions, Draw photon from
spectrum
Time t
GRBSimulation
SpectObj
24
Conclusion

• GRB package up and running
• Physical description of Fireball model coded and
  embedded into Gleam.
• Possibility to process the spectrum that comes
  from an astrophysical source model
• Possibility to implement the physics and see the
  observational capability of GLAST
• Currently redesigned for improved
  clarity/extension
• More scenarious
• More object oriented
• More User Friendly
• Extension to low energy (GBM integration !!)
• Goodies to come and somehow related to the GRB
  physical model
• Physical model fitting engine (A10)
• Give an estimation of the goodness of fit
• Parameter estimation
• Constrain on the model adopt
• Quantum Gravity effect Flux(E) Flux(E(z))
• Disentangle all the known physical effects
  (related to the time lag) to isolate the effect
  of QG
• Include the dispersion law preview by the theory
  to add the effect of QG

25
The problem of the Interval

• In the general definition of the Spectrum object
  (see ISpectrum.h) the interval to wait before the
  next event is 1/Rate.
• For transient sources the Rate depends on the
  time

Rate
Time

• The solution we adopt is compute the integral of
  1/Rate