Preliminary Profile Reconstruction of EA Hybrid Showers

1
Preliminary Profile Reconstruction of EA Hybrid
Showers

• Bruce Dawson Luis Prado Jr
• thanks to Brian Fick Paul Sommersand Stefano
  Argiro Andrea de Capoa

Malargue, 23 April 2002
2
Introduction

• we are using
• the Flores framework
• hybrid geometries from Brian and Paul
• profile reconstruction scheme described
  inGAP-2001-16
• absolute calibration derived from remote laser
  shots GAP-2002-10
• profiles viewable (December - March) at
  www.physics.adelaide.edu.au/bdawson/profile.htm

3
Basic Steps

• determine light collected at the detector per 100
  ns time bin
• F(t) (units 370nm-equivalent photons at
  diaphragm)
• determine fluorescence light emitted at the track
  per grammage interval
• L(X) (units of photons in 16 wavelength bins)
• requires subtraction of Cherenkov contamination
• determine charged particle number per grammage
  interval
• S(X) (longitudinal profile)

4
(No Transcript)
5
(No Transcript)
6
Received Light Flux vs time, F(t)

• Aim to combine signal from all pixels seeing
  shower during a given 100ns time slice
• Avoid including too much night sky background
  light
• Take advantage of good optics
• good light collection efficiency
• try (first) to avoid assumptions about light spot
  size (intrinsic shower width, scattering)
• variable c method developed to maximize S/N in
  flux estimate

7
Light Flux at Camera F(t) (cont.)

• assume track geometry and sky noise measurement
• for every 100ns time bin include signal from
  pixels with centres within c of spot centre.
• Try values of c from 0o to 4o. Maximize S/N over
  entire track

8
Optimum Chi values
9
Camera - Light Collection
10
Event 33 Run 281 (bay 4) January
11
Longitudinal Profile S(X)
Received LightF(t)

• First guess, assumes
• light is emitted isotropically from axis
• light is proportional to S(X) at depth X
• True for fluorescence light, not Cherenkov light!

shower geometry,atmospheric model
map t onto slant depth X
Light emitted at track L(X)
fluorescence efficiency
Shower size at track, S(X)
12
Complications - Cherenkov correction

• Cherenkov light
• intense beam, directed close to shower axis
• intensity of beam at depth X depends on shower
  history
• can contribute to measured light if FD views
  close to shower axis (direct) or if Cherenkov
  light is scattered in direction of detector

13
This particular event
Event 33, run 281 (bay 4), December
Rp 7.3km, core distance 11.8 km, theta 51
degrees
14
Cherenkov correction (cont.)

• Iterative procedure

15
Smax
number of iterations
16
Xmax
number of iterations
17
(No Transcript)
18
Finally, the profile S(X)

• this Cherenkov subtraction iteration converges
  for most events
• transform one final time from F(t) to L(X) and
  S(X) using a parametrization of the fluorescence
  yield (depends on r, T and shower age, s)
• can then extract a peak shower size by several
  methods - we fit a Gaisser-Hillas function with
  fixed Xo0 and l70 g/cm2.

19
E2.5x1018eV, Smax1.8x109, Xmax 650g/cm2
particle number
atmospheric depth (g/cm2)
20
Energy and Depth of Maximum

• Gaisser-Hillas function
• Fit this function, and integrate to get an
  estimate of energy deposition in the atmosphere
• Apply correction to take account of missing
  energy, carried by high energy muons and
  neutrinos (from simulations).

21
Missing energy correction
Ecal calorimetric energyE0 true
energy from C.Song et al. Astropart Phys (2000)
22
Event 336 Run 236 (bay 4) December
Rp 10.8km, core distance 11.1 km, theta 26
degrees
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
Event 336 Run 236 (bay 4) December
photons (equiv 370nm)
time (100ns bins)
27
E 1.3 x 1019eV, Smax 9.2 x 109, Xmax
670g/cm2
particle number
atmospheric depth (g/cm2)
28
Event 751 Run 344 (bay 5) March
photons (equiv 370nm)
time (100ns bins)
29
Comparison of two methods
30
E 1.5 x 1019eV, Smax 1.0 x 1010, Xmax
746g/cm2
particle number
atmospheric depth (g/cm2)
31
Shower profile - two methods
32
2 Methods Compare Nmax
33
Events with bracketed Xmax

• 57 total events
• (all bay 4 hybrid events six bay 5 hybrid
  events from March)
• of these 35 had reasonable profiles where Xmax
  appeared to be bracketed (or close to).

34
Nmax distribution
35
Shower Energy
36
Shower Energy dN/dlogE
37
Xmax distribution
38
(No Transcript)
39
(No Transcript)
40
(No Transcript)
41
Conclusions

• First analysis of hybrid profiles is encouraging,
  with some beautiful events and the expected
  near-threshold ratty ones
• preliminary checks with alternative analysis
  methods indicate that we are not too far wrong in
  our Nmax assignments
• we are continuing our work to check and improve
  algorithms

42
(No Transcript)