E-165 FLASH Measurement of Air Fluorescence Produced by Air Showers.

1
E-165FLASHMeasurement of Air Fluorescence
Produced by Air Showers.
2
E-165
Fluorescence from Air in Showers (FLASH) J.
Belz1, Z. Cao2, P. Chen3, C. Field3, P.
Huentemeyer2, W-Y. P. Hwang4, R. Iverson3,
C.C.H. Jui2, T. Kamae3, G.-L. Lin4, E.C. Loh2,
K. Martens2, J.N. Matthews2, W.R. Nelson3, J.S.T.
Ng3, A. Odian3, K. Reil3, J.D. Smith2, P.
Sokolsky2, R.W. Springer2, S.B. Thomas2, G.B.
Thomson5, D. Walz3 1University of Montana,
Missoula, Montana 2University of Utah, Salt Lake
City, Utah 3Stanford Linear Accelerator Center,
Stanford University, CA 4Center for Cosmology and
Particle Astrophysics (CosPA), Taiwan 5Rutgers
University, Piscataway, New Jersey
Collaboration Spokespersons
3

• Cosmic Rays have been observed with energies at
  up to 1020 eV
• The flux (events per unit area per unit time)
  follows roughly a power law
• E-3
• Changes of power-law index at knee and ankle.
  
• Onset of different origins/compositions?
• Where does the spectrum stop?

4
Greisen-Kuzmin-Zatsepin (GZK) Cut-off
31020 eV
50 Mpc Size of local cluster
(protons)

• Protons above 61019 eV will loose sizable energy
  through CMB
• Super-GZK events have been found with no
  identifiable local sources

5
Extensive Air Showers
Zoom on next slide
6
UHECR From Source to Detector
CMB ?
7
Observation of Cosmic Ray with Fluorescence
Technique

• The two detector sites are located 12 km apart
• Geometry of an air shower is determined by
  triangulation.
• Energy of primary cosmic ray calculated from
  amount of light collected.

8
(No Transcript)
9
UHECR Spectrum AGASA vs HiRes
10
(No Transcript)
11
Fetish - style of the Atie/Atye/Attie/Attye from
Ivory Coast

• Fetish (feitico), another reprehensible term
  introduced into West Africa by the Portuguese, is
  wrongly used to describe the minor divinities.
  Fetishism and juju are originally derogatory
  terms and are relevant to nothing of African
  Religion.

12
Energy Fetishism

• Fetish - "something evoking irrational devotion
  or respect".
• Measure photons or charged particle densities.
• Energy - distant derivation.
• There are many linear and non-linear steps.
• Fluorescence efficiency is primary.

13
Fluorescence efficiency is the foundation for our
belief that we are measuring energy

• How well is it known ( for an ionizing particle)?
• Is it linear with particle number (size) in an
  extensive air shower?
• Can it be affected by accidental conditions?
  Impurities, etc.
• How do we determine answers to these issues with
  sufficient accuracy?

14
Second Knee, showing correlation between knee
energy and spectral normalization
15
Second Knee Spectrum, Shifted to make knee come
out at same energy
16
Second Knee, cont.

• All experiments agree when a scale shift is
  applied.
• But what is the actual energy of the second knee?
• Fluorescence method should be very reliable (
  nearby events, little atmospheric attenuation ).
• Position comes primarily from our knowledge of
  air fluorescence efficiency.

17
FLASH useful for future UHECR Experiments
Ground-Based The Pierre Auger Observatory

• Hybrid detection
• 1600 Cherenkov detectors 1.5 km gridin 3000 km2
• 4 fluorescence eyes Comparable to HiRes

65 km
18

• Space-Based EUSO, OWL/AirWatch

19
Bunner Air-Fluorescence Spectrum
20
Current Understanding

• Bunner (1967), Kakimoto et al. (1995), Nagano et
  al. ( 2002, unpublished) indicates 15
  systematic errors in overall yield and larger
  errors in individual spectral lines.
• Ground based experiments non-linear effects
  possible due to ? dependence of atmospheric
  attenuation.
• At 30 km, event energy can change by 25 if 390
  nm line intensity changes by 40.

21
Photon yields between 300 and 406nm from Nagano,
Kakimoto( HiRes) and Bunner
22
Comparison of Fluorescence Yields for major
spectral line groups
23
Relative Contributions of Different Spectral
Lines at Different Horizontal Distances
24
Importance to EUSO and OWL (space-based)
experiments

• Path-lengths from shower to detector almost
  constant small ? dependence (10 integral
  variation over different fluorescence models)
• Most showers detected over oceans effect of
  water-vapor and other impurities may be
  important. Some evidence for H2O quenching
  already exists.

25
Fluorescence Pressure Dependence
t lifetime
Ppressure
Ttemp.
26
Reference Pressure
27
Fluorescence Yield

• Y photons per meter-ionizing particle
• P, P pressure and reference pressure
• C spectral line normalization

28
Example of Dependence of Air Fluorescence from
Nagano et al (personal communication) at .85 MeV
29
Life time vs Pressure (Air) from Nagano et
al.(p.c.)-at .85 MeV
30
Life time vs Pressure (Air) from Nagano et
al.(p.c.)-at .85 MeV
31
Life time vs Pressure (Air) from Nagano et
al.(p.c.)-at .85 MeV

• Air fluorescence lifetimes 25 nsec
• Bunner quotes lifetimes near 40 nsec
• Large uncertainties remain.

32
Dependence on electron energy ( Kakimoto et
al.NIM, 1995)
33
Dependence on electron energy ( Kakimoto et
al.NIM, 1995)
34
SLAC test run results

• Two week run in June 2002
• Prototype thin target setup
• Measured pressure dependence integrated over
  300-400 nm.
• Measured average lifetime over 300-400 nm.
• Confirmed linear behavior of Y with respect to
  beam current below 109 ppb

35
T461 Setup
LEDs
PMTs
36
SLAC test result on linearity
37

• Effect was enhanced at low pressures and was
  reduced near sea-level pressure

38
Simple model for non-linearity

• Beam pulse passes in 3 picoseconds
• Treat pulse as a uniform ball of charge and
  estimate impulse ?p it imparts on an electron
  (and hence the kinetic energy gained ?EK
• ?p Ne2/4pe0r2 ? ?t (r 1 mm)
• ?EK (?p)2/2m 700 eV for N1010 IF the
  ejected electron does not suffer collisions
  while being acceleratedthis is enough energy to
  cause secondary ionization
• Important quantity here is the Mean Free Path ?.

39
Mean Free Path

• At room temperature in air/nitrogen, we have
• ? 2x10-4/P
• ? in meter,
• P in torr

40
Mean Energy Gained(before collision)

• Estimate the energy gained by an electron before
  collision
• ?EK Ne2/4pe0
• 1/r - 1/(r?)

41
SLAC test result- comparison of N2 and Air
efficiency
42
Ne2x109
43
N2 Decay Time Measurement
44
SLAC test N2 Decay Time Measurement
45
N2 (in air) Decay Time Measurement
46
Dependence on electron energy Kakimoto et al.
and T461
47
SLAC Test and Beyond

• Test clearly established ability to detect air
  fluorescence in FFTB beams.
• Test showed that we can measure the pressure
  dependence and fluorescence lifetime integrated
  over total spectrum
• What is needed, however, is spectrally resolved
  pressure and lifetime measurement.
• Test only measured Y at 28.5 GeV. Energy
  dependence over realistic shower energies is
  required.

48
OBJECTIVES OF E-165

• Spectrally resolved measurement of fluorescence
  yield to better than 10.
• Investigate dependence on electron energy.
• Study effects of atmospheric impurities.
• Observe showering of electron pulses in air
  equivalent substance (Al2O3) with energy
  equivalents around 1018 eV.

49
Proposed Program

• Gas Composition
• N2/O2 dependence, and Ar, CO2, H2O impurities
• Pressure Dependence
• Yield versus Pressure down to 10 torr
• Energy Dependence
• Yield versus electron energy distribution down to
  100keV
• Fluorescence Spectrum
• Resolve individual bands using narrow band
  filters or spectrometer.
• Pulse Width
• Pressure dependence of fluorescence decay time
  for each spectral band

50
THIN TARGET STAGE

• Pass electron beam through a thin-windowed air
  chamber.
• Measure the total fluorescence yield in air at
  30 GeV.
• Measure the yield over wide range of pressures at
  and below atmospheric.
• Measure emission spectrum using narrow band
  filters or spectrometer.
• Effects of N2 concentration. Pure N2 to air. Also
  H2O, CO2, Ar, etc.

51
(No Transcript)
52
T461 Setup
LEDs
Beam
Gas Flow
PMT
53
General Requirements for thin target run

• FFTB downstream of last magnets
• lt1 R.L. in beam no radiation problem
• Require 108 109 e/pulse for linear operation.
• Require improved toroid sensitivity to monitor
  beam at this intensity (or equivalent
  cross-calibrated measurement).

54
THICK TARGET STAGE

• Pass electron beam through varying amounts of
  air equivalent showering material (Al2O3).
• Measure light yield as a function of depth in the
  shower ( sample light from a wide range of
  electron energies).
• Is fluorescence proportional to dE/dx?
• What are the contributions of low-energy
• (lt1 MeV) electrons?
• Can existing shower models (EGS, GEANT, CORSIKA)
  correctly predict fluorescence light?
• How does the fluorescence yield in an air shower
  track the shower development?

55
Comparison- Cosmic/SLAC

• Cosmic Ray
• Dump 1017 to 1020 eV per particle into atmosphere
• 600-800 gm/cm2 into shower ( Xmax) particles are
  electrons with energies between 100 keV and few
  GeV.

• SLAC beam
• Dump 3 x 1018 - 3 x 1019 eV per beam bunch into
  Alumina target
• 200 gm/cm2 (Xmax), particles are electrons with
  energies between 100 keV and 2 GeV.

56
(No Transcript)
57
THICK TARGET SETUP
58
CORSIKA AIR SHOWERS
59
BREMSSTRAHLUNG BEAM OPTION
?
60
THICK TARGET SHOWER DEVELOPMENT
61
(No Transcript)
62
Thick Target Requirements

• Electron shower implies x 100 multiplier
• Can run beam intensity 107 to 108 (radiation OK)
  with same signal strength (or brem. option).
• Showering beam spreads out in air (50 cm)
  Careful calculation/measurement of optical
  acceptance necessary.
• Measurement with optical masks to check relative
  tube response to radial displacement of source
  during run.
• LED mapping of optical response off-line.

63
SYSTEMATIC UNCERTAINTIES

• Beam charge should be measurable by the beam
  toroids to better than 2.
• The uncertainties in showering 3.
• Detector systematic uncertainties of 5.4.
• Detector Optics 4 (thin) 6.5 (thick).
• Total systematic uncertainty of 7-9.

64
SYSTEMATIC UNCERTAINTIES
Thin Target Thick Target
Beam 2 2.2
Showering - 3
Detector System 5.4 5.4
Optical System 4 6.5
Total 7 9.2
65
T461 PMT Stability (2.2).
66
SYSTEMATIC UNCERTAINTIES

• Beam charge should be measurable by the beam
  toroids to better than 2.
• When showering the beam, the beam energy will
  also affect the number of particles in the
  shower. This should be determined to better than
  0.5.
• If a bremsstrahlung beam is used the contribution
  of the converter foil thickness uncertainty
  should be less than 1.

67
T461 Beam Stability (2)
x 1010 e-
68
SYSTEMATIC UNCERTAINTIES

• The uncertainties in showering 3.
• Uncertainty in simulations and transition effects
  from dense target to air 2.
• Uncertainty in amount of showering material of
  1-1.5.

69
SYSTEMATIC UNCERTAINTIES

• Detector systematic uncertainties of 5.4.
• PMT calibration uncertainty of 5.
• Cable and ADC uncertainty of 2.
• Detector Optics 4 (thin) 6.5 (thick).
• Wide band filters and mirrors (1).
• Narrow band filter transmission (3).

70
CONCLUSION

• FLASH aims to achieve an accuracy of 10 in the
  total fluorescence yield and individual spectral
  lines.
• Verify energy dependence of yield down to
  100keV.
• Both thin target and thick target approaches will
  be invoked.
• Dependence of yield and spectrum on pressure and
  atmospheric impurities will be measured.
• Shower developments equivalent to 1018 eV will
  be measured at various depths and compared with
  codes.
• We hope that FLASH will help to shed light on the
  apparent differences between HiRes and AGASA, and
  provide reliable information for future
  fluorescence-based UHECR experiments.

71
Beyond FLASH?

• Suppose thick target shows discrepancy.
• Is the problem due to fluorescence yield or
  shower modeling?
• Need for thin-target low-energy studies
• Orion beam energies will be of great interest
  but need to go to lower energies
• Ideally, from 100 keV up. Can use absorbers to
  look at shower decrement

72
Dependence on electron energy Kakimoto et al.
and T461
-----