ESAF atmosphere simulation

1
ESAF atmosphere simulation

• status and progress

2
Overview

• Single varying parameter productions, some
  studies in progress
• shift of fluorescence profile w.r.t. electrons
  profile
• Fluo photons time distribution on pupil (nb and
  shape)
• Single scattered Cerenkov contribution to signal
• a word on ground detector
• Reflected Cerenkov peak
• MCRadiativeTransfer, a Monte-Carlo based
  simulation of multiple scattering
• principle
• response to a spot-like source of light
• Comparison with BunchRadiativeTansfer
• some examples

3
Light production ingredients

• integration over electrons energy spectrum
• parametrized Giller, Nerling
• single energy
• Local atmospheric conditions taken into account
• Several fluorescence modules
• Kakimoto, Nagano measurements
• theoretical yield (in progress)
• Analytical cerenkov generation within 300-450
  nm

4
Yield studies
Fluo

• Fluorescence
• Weak dependence with altitude for any model
• single energy showers are good approximations
• Cerenkov
• single energy 80MeV largely overestimates the
  yield
• single energy 24MeV without threshold effect is a
  good approximation

cerenkov
5
Shift of fluorescence profile w.r.t electrons
6
Light profiles at creationShift w.r.t. electrons
one
Fluo
elec
E 1020 eV X1 within 5,100 g/cm2 USStd
atmosphere Nadir impact

• Fluorescence profile Hmax is shifted w.r.t.
  electrons one 500 meters
• Not Cerenkov (20 - 50 m)

cerenkov
elec
7
Xmax shift

• shift for fluorescence - 60 g/cm2
• for cerenkov - 6 g/cm2

more electrons but shorter path
less electrons but longer path
?L
?X
shower Xmax
fluo Xmax
Crossed air depth
8
RadiativeTransfer ingredients

• Lowtran7 (rayleigh, ozone, aerosols) used for
  lambda dependent transmission calculations
• photons gathered into bunches, then propagated
• Lateral and angular distributions are used for
  correction in position
• In following productions, only clear sky
  conditions rayleigh and ozone effects

9
Photons on pupil General features
Theta fixed 60 deg
HALF Fluo
scattered ckov
X1 within 5,100 g/cm2 USStd atmosphere Nadir
impact
refelcted ckov

• Half fluorescence integrated until maximum
• reflected

Energy fixed 1020 eV
10
Shape of fluorescence time distribution on pupil
11
Shape of Fluo timedistribution on pupil
Half Nph

• Gaussian fit applied to fluorescence time profile
  on pupil -gt Nmax and Width

Nmax
Width (GTU)
Nmax / (Half Nph)
12
Signal width
Width (GTU)

• Width Hmax correlations

E 1020 eV X1 within 5,100 g/cm2 USStd
atmosphere ?local 60? , ? 0?
a transversal track is narrower in time
Y (km)
shower direction
X (km)
13
Backscattered cerenkov contribution to signal
14
Scattered cerenkov contribution

• Air backscattered cerenkov (scattered
  cerenkov) / fluo ratio is constant with theta
• but rayleigh phase function makes it depend on
  position in FoV

Y (km)
shower direction
E 1020 eV X1 within 5,100 g/cm2 USStd
atmosphere ?local 60? , ? 0?
X (km)
15
Comparison with Ground detectorCerenkov pollution

• EUSO looking vertically upward
• 1019 eV, ? 60?, ? 0?
• no cut on photon incident direction
• both scattered and direct cerenkov contributions
• when shower does not point toward detector, it
  seems cerenkov contribution quite feable (10-15)
• but do not forget aerosols, fluorescence
  scattering contributions

16
Some examples of profiles
1
1
5 ?s large
3
2
100 ?s large
60 ?s large
17
Reflected Cerenkov

• lambertian surface
• albedo 5
• angular distribution -gt spread effect

18
Reflected cerenkov

• Nadir shower only on these plots, albedo 5
• Using mean detector effect, photons are converted
  into photo-electrons, then a signal-to-noise
  ratio is defined ask to be gt 2 (cf. red book)
• blue old background
• red new one
• At high theta, lot of atmosphere to travel before
  ground

Theta
19
Reflected cerenkov Fov effect
Nph
Y (km)
shower direction

• only showers with ? 60 deg
• lambertian surface,albedo 5
• a dissymmetry appears effect of earth
  sphericity on cerenkov spread

X (km)
Nmax
Y (km)
shower direction
X (km)
20
Effects of aerosols on reflected cerenkovSome
examples
rural, 23 km vis.
rural, 5 km vis.
Transmission bunch absorption
Transmission bunch transparent
21
a new algorithm for Radiative Transfer
22
The principle

• No photons will be scattered closer than Dmin
• Corresponding solid angle value ?max is the same
  for all the photons
• Poisson(Nph ?max) gives the nb of photons to
  propagate N?
• Photons can be absorbed or go out of atmosphere
• Ozone in atmospheric low layers is not taken into
  account

Dmin (? ?max)
23
some precisions

• Each scattering order must be simulated
  separately, so need to recreate N? photons each
  time
• To save CPU time, ?max is replaced by the solid
  angle value at creation (a posteriori checks show
  it is reasonable)
• Cerenkov and fluorescence scattering are
  simulated
• So far, only clear sky conditions supported
  Lowtran is not used for scattering processes, but
  kept for last transmission to detector (ozone
  component)
• In optimized mode, CPU time for first order 1
  min

24
Response to a spot-like isotropic source
25
Detailed contribution of each order

• simulated until the fifth order included
• a decrease of a factor of 2 from an order to the
  next one
• a delay in time

26
Response to a spot-like focused source
27
Detailed contribution of each order

• simulated until the fifth order included
• no direct light
• first order corresponds to reflected and
  backscattered usual component
• a decrease of a factor of 2 from an order to the
  next one (excepted for the first one)
• a delay in time

28
A typical shower 1020, 60 deg
29
(No Transcript)
30
(No Transcript)
31
Is multiple scattering polluting the signal ?

• A signal-to-noise ratio is defined at each
  GTU(SNR gt 1 with background0.4)
• Scattered fluorescence seems not pollute
  significantly. Cerenkov does.

32

• horizontal shower (87 deg)
• 1020 eV
• only first order

33
(No Transcript)
34
(No Transcript)
35
Cross-check with BunchRadiativeTransfer

• Same mean distribution for single scattered and
  reflected cerenkov (error lt 5)
• fluctuations are the same

36
(No Transcript)
37
Gaussian fit v.s. Hmax
38
Cerenkov substraction

• Profile total en noir, profile fluo seule en
  rouge
• A ? gt 60 deg, correspondant aux altitudes gt 8 km,
  le cerenkov diffuse est spontanement soustrait
  par le fit gaussien
• Un procede de fit en deux temps donne des
  evolutions en Hmax plus proches de la fluo seule,
  mais introduit des fluctuations

?
Prochaine etape reconstruction
Hmax
39
Cerenkov peak dissymmetry illustration
1
1
2
3
150 ?s de large
3
2
200 ?s de large
100 ?s de large
40
(No Transcript)