Transcript A Measurement of the Ultra-High Cosmic Ray Flux with the HiRes

The UHECR Spectrum
observed with HiRes in
monocular mode
Andreas Zech
(LPNHE, Paris)
Seminar at UNM
Albuquerque, 03/29/05
Outline
• Ultra-High Energy Cosmic Ray Physics
• The HiRes Experiment
• Unfolding the Cosmic Ray Spectrum
• Fits to the Spectrum
• Summary
• The Future of HiRes: TA & TALE
Ultra-High Energy
Cosmic Ray Physics
Energy Spectrum
• differential flux:
dN / (dE A  dt)
• follows roughly
E-3 power law
• direct observation not
possible above 1 PeV
• two widely observed
features:
– ‘knee’ at ~1015.5 eV
– ‘ankle’ at ~1018.5 eV
knee
second knee
ankle
Propagation Effects
• magnetic fields (galactic, extragalactic)
• red-shifting
• e+e-- pair production with CMBR (at ~ 1017.8 eV)
• photo-spallation of cosmic ray nuclei
• GZK effect with CMBR (at ~ 1019.8 eV)
 (2.7 K) + p
(1232)
+ + n
 (2.7 K) + p
(1232)
o + p
Strong flux suppression expected for extra-galactic sources.
Extensive Air Showers
main channels:
+(-)
+(-) +  (
 )
o
2
K+(-)
+(-) +  (
 )

2
main e.m. processes:
• bremsstrahlung
Ground Arrays (Surface Detectors)
• Detection of lateral particle
profile on ground.
• Reconstruction of geometry
from pulse & time information.
• Reconstruction of energy by
model comparisons.
AGASA
• Pro: 100 % duty cycle, low cost,
low maintenance, good geometry
reconstr., nearly constant aperture
• Contra: Energy reconstr. is model
dependent, uncertainties due to
fluctuations in lateral profile.
Air Fluorescence Detectors
• Detection of longitudinal
shower profile via UV
fluorescence light.
• Reconstruction of geometry
from recorded shower ‘track’.
• Using the atmosphere as a
calorimeter.
Fly’s Eye
• Pro: Direct measurement of cosmic ray
energy and shower maximum, good
geometry & energy reconstruction.
• Contra: 10 % duty cycle, higher cost &
maintenance, energy dependent
aperture, atmospheric uncertainties
UHECR Composition
• depth of shower
maximum ( Xmax )
depends on energy &
cosmic ray species
=> indirect composition
measurement
• comparison of Xmax
with simulation
allows bi-modal
determination of c.r.
composition in a
statistical way.
The HiRes
(High Resolution Fly’s Eye)
Experiment
The HiRes Collaboration
J.A. Bellido, R.W. Clay, B.R. Dawson,
K.M. Simpson
N. Manago, M. Sasaki
University of Tokyo
University of Adelaide
J. Boyer, S. Benzvi, B. Connolly,
Finley, B. Knapp, E.J. Mannel,
O’Neil, M. Seman, S. Westerhoff
Columbia University
J. Belz, M. Munro, M. Schindel
Montana State University
G. Martin, J.A.J. Matthews, M. Roberts
C.
A.
T. Abu-Zayyad, J. Albretson, G. Archbold,
J. Balling, K. Belov, Z. Cao, M. Dalton,
A. Everett, J. Girard, R. Gray, W. Hanlon,
P.
Hüntemeyer, C.C.H. Jui, D. Kieda,
K.
Kim, E.C. Loh, K. Martens,
J.N.
Matthews, A. McAllister, J. Meyer, S.A.
Moore, P. Morrison, J.R. Mumford,
K. Reil,R. Riehle, P. Shen, J. Smith,
P.
Sokolsky, R.W. Springer, J. Steck,
B.T.
Stokes, S.B. Thomas,
T.D.
Vanderveen, L. Wiencke
University of Utah
University of New Mexico
D. Bergman, L. Perera, G. Hughes,
S. Stratton, D. Ivanov,
S. Schnetzer, G.B. Thomson, A. Zech
Rutgers University
J. Amann, C. Hoffman, M. Holzscheiter,
Marek, C. Painter, J. Sarracino,
Sinnis, N. Thompson, D. Tupa
L.
G.
Los Alamos National Laboratory
HiRes-1 consists of one ring
of 22 mirrors. Coverage in
elevation is from 3 to 17 deg.
Sample & Hold Electronics
are used to record pulses.
(5.6 µs window)
HiRes-2 has two rings of 21 mirrors
each. Coverage in elevation from 3
to 31 deg.
Flash ADC electronics record signals
at a frequency of 10 MHz.
• Mirror area ~ 5 m2 .
• 256 (16x16) PMT per mirror.
• One PMT sees ~ 1 degree of the sky.
Measuring the Energy Spectrum with HiRes
Stereo observation of the
cosmic ray flux yields a
better resolution in
geometry and energy
than monocular.
Analyzing our data in
monocular mode has also
some advantages, though:
• better statistics at the high
energy end due to longer
lifetime of HiRes-1.
• extension of the spectrum to
lower energies due to greater
elevation coverage and better
time resolution of HiRes-2.
1. Reconstruction of the shower-detector plane
• project signal tubes onto sky
• fit tube positions to a plane through
the center of the detector
• reject tubes that are off-track
off in time) as noise
=> shower axis lies in the fitted
shower-detector plane
(and
2. Reconstruction of the
geometry within the
shower-detector-plane
3. Shower Profile & Energy
Reconstruction
• Reconstruct charged particle
profile from recorded p.e.’s .
• Fit profile to G.H. function.
• Subtract Čerenkov light.
• Multiply by mean energy loss rate
 => calorimetric energy
• Add ‘missing energy’ (muons,
neutrinos, nuclear excitations;
~10%) => total energy
Phototube Calibration
• Absolute calibration
using a portable lightsource (“RXF”), that
is carried to both sites
about once a month.
– calibration of RXF in
the lab using HPDs.
=> +/- 10% uncertainty
in energy scale.
• Relative calibration
at the beginning and
end of each nightly
run.
– using YAG laser
– optical fibers distribute
the laser signal to all
mirrors.
Atmospheric Calibration
• Rayleigh contribution is
quite stable and well known.
• Detailed monitoring with
steerable lasers at both sites.
• Aerosol profile of the
atmosphere has to be
monitored during the run.
• Additional vertical laser
outside of Dugway (Terra).
• “Shoot the Shower”
=> <VAOD> = 0.04 +/- 0.02
=> +/- 15 % in J(E)
Unfolding the Cosmic Ray
Spectrum
Deconvolution of the UHECR Spectrum
We observe the spectrum convoluted with detector acceptance
and limited resolution.
Deconvolution with help of a correction factor:
D(Ei)= Rij T(Ej)
T(Ei)= [Gmc(Ei)/Rmc(Ei)] D(Ei)
We need M.C. to simulate acceptance (& resolution)
of our detectors for the flux measurement:
This requires a simulation program that describes the shower
development and detector response as realistically as
possible.
HiRes Monte Carlo Simulation
CORSIKA Shower Library (proton & iron)
Fit parameters scale with
primary energy:
Gaisser-Hillas fit to the shower
profile:
Data / Monte Carlo Comparisons
Testing how well we understand and
simulate our experiment...
• HiRes-1:
– data shown from 06/1997 to 02/2003.
– 6920 events in final event sample
• HiRes-2:
– data shown from 12/1999 until 09/2001.
– 2685 events in final event sample
• Measurement of average atmosphere used
• M.C. : ~ 5 x data statistics
HiRes-2: light (# p.e. / deg of track)
HiRes 2: 2/d.o.f. of time vs. angle fit
Energy Distribution & Resolution
 =18%
HiRes-1: distance to shower core
HiRes-1: Energy Resolution
Instant Apertures
HiRes-1
HiRes-2
The HiRes-2 UHECR Spectrum
HiRes and Fly’s Eye
HiRes and Haverah Park
HiRes and Yakutsk
HiRes and AGASA
Systematic Uncertainties
Systematic uncertainties in the energy scale:
•
•
•
•
absolute calibration of phototubes: +/- 10 %
fluorescence yield: +/- 10 %
correction for ‘misssing’ energy: +/- 5 %
aerosol concentration: ~ 9 %
=> uncertainty in energy scale: +/- 17 %
+ atmospheric uncertainty in aperture
=> total uncertainty in the flux: +/- 31 %
Systematics due to MC Input Composition
• Detector acceptance at
low energies depends on
c.r. composition.
• MC uses HiRes/MIA
measurement as input
composition.
• Relevant uncertainties :
– detector calibration
– atmosphere
– fit to HiRes/MIA data
=> +/-5 % uncertainty
in proton fraction
Systematics due to Atmospheric Variations
• Repeated HiRes-2
analysis using the
atmospheric database.
• Regular Analysis:
– <HAL>=25 km,
<VAOD>=0.04
– in MC generation
– in data & MC reconstr.
• Systematics Check:
– HAL & VAOD from
database (hourly entries)
– in MC generation
– in data & MC reconstr.
spectrum with database
spectrum with average
Fits to the Spectrum
Power Law Fits: Observation of Ankle
and Evidence for High Energy Break
• fit without break points:
2 / d.o.f = 114 / 37
• fit with one break point:
2 / d.o.f. = 46.0 /35,
logE=18.45+/-0.03 eV
• fit with two break points:
2 / d.o.f. = 30.1 / 33,
logE=18.47+/-0.06 eV
&
19.79+/-0.09eV
=3.32+/-0.04 & 2.86+/0.04 & 5.2+/-1.3
• In case of unchanged
spectrum above 2nd break
point, we’d expect 28.0
events where we see 11
=> Poisson prob.: 2.4 E-4
Fit with Toy Model
• Fit to the HiRes monocular
spectra assuming
– galactic & extragalactic
components
– all propagation effects
(e+e-, red-shift, GZK)
• Details of the fit procedure
– Float normalization, input
spectral slope () and m
– uniform source density
evolving with (1+z) m
– Extragalactic component
• 45% protons at 1017 eV
• 80% protons at 1017.85 eV
• 100% protons at 1020 eV
– Use binned maximum
likelihood method
Extragalactic
Galactic
Interpretation
• Pion-production pileup
causes the bump at
1019.5 eV.
• e+e- pair production
excavates the ankle.
• Fractionation in
distance and energy;
e.g., z=1 dominates
at second knee.
The Future of HiRes: TA / TALE
TA - the “Telescope Array”
• SD: 576 scintillation counters,
each 3 m2 area, 1.2 km spacing.
• 3 fluorescence stations, each
covering 108o in azimuth,
looking inward.
• Central laser facility.
• Millard County, Utah, flat
valley floor for SD, hills for
fluorescence, low aerosols.
• A 1020 eV event (on a night
when the moon is down) will be
seen by SD and all three
fluorescence detectors.
• A powerful detector for hybrid
and stereo cross correlation
with SD.
Ideas for Recyling HiRes
• Two HiRes detectors, moved to
Millard Co.
• 6 km stereo with TA
fluorescence detectors.
• Each HiRes detector has two
rings, 270o azimuthal coverage.
• Aperture of 16000 km2 ster.
• Increase fluorescence aperture
from 500 to 1,780 km2 ster,
including 10% duty cycle. (TA
SD=1400).
• Increase in fluorescence
aperture of x 3.6
TA Low energy Extension:
“Tower of Power” & Infill Array
• 15 mirrors, 3xHiRes area,
in rings 3,4,5 ( 3o - 71o )
=> good coverage down to
logE = 16.5 eV
• 111 AGASA counters,
spacing of 400m, shown
in red.
• 10 x HiRes/MIA hybrid
aperture.
=> observation of spectrum
& composition around
second knee
for more information:
www.cosmic-ray.org
www.physics.rutgers.edu/~aszech
Fit with Toy Model
• Fit to the HiRes monocular
spectra assuming
– galactic & extragalactic
components
– all propagation effects
(e+e-, red-shift, GZK)
 = 2.32+/-0.01
• Details of the fit procedure
– Float normalization, input
spectral slope () and m
– uniform source density
evolving with (1+z)3
– Extragalactic component
• 45% protons at 1017 eV
• 80% protons at 1017.85 eV
• 100% protons at 1020 eV
– Use binned maximum
likelihood method
Extragalactic
Galactic
Summary
• We have measured the UHECR spectrum from
1017.2 eV to the highest energies with the HiRes
detectors in monocular mode.
• A simulation of the exact data taking conditions was
used to determine the acceptance and resolution of the
detector, and tested in detail against data.
• We observe the ‘ankle’ in the HiRes-2 spectrum
at 1018.5 eV.
• The combined monocular HiRes spectra show evidence
for a break above 1019.8 eV. The Poisson probability for
continuation of the spectrum with unchanged slope
from the HiRes monocular data is 2.4 * 10-4 .
Cosmology with
TA/TALE ?
• Adjust evolution to match
QSO’s:
• m=2.6, z<1.6
• Lower m, z>1.6
• Must extend spectrum
measurement lower by an
order of magnitude.
Mono versus Stereo Energy Measurements
HiRes-1 mono
vs. stereo
The HiRes
monocular
energy is in
excellent
agreement with
stereoscopic
measurements !
Calibration Correction
• Problems with the
HiRes-2 calibration due
to limited access to
Dugway.
• We adopted HiRes-1
calibration for the
absolute energy scale.
• Correction factors for
each dataset were
determined from
comparisons of
stereo events.
-22 %
-5 %
-11 %
<-14 %>
Varying Detection Parameters
• Trigger logic
=> data divided into 3
datasets
• Weather
=> strict cuts based on
hourly observation
• Trigger gains
• Dead mirrors
• Live-time
=> Nightly Database
• Aerosols
=> atmospheric database
from laser shots
=> average values were used
for this analysis
• Atmospheric Density
=> Seasonal variations
• Light pollution
=> Average for each data set
Noise assisted triggering
Track angle distribution
shows a deficit in the
MC for nearly
vertical tracks.
Noise assisted triggering
Adding noise to the
MC increases the
number of nearly
vertical tracks.
This effect is caused
by an inefficiency in
the HiRes-2 trigger.
Ambient noise (low amplitude) is
added to each channel in the MC.
It is measured from the variances
taken from “snapshots”.
Additional sky noise
(high amplitude)
is added to the M.C. to
get agreement with
data of a certain period.
Fits to the HiRes-2 Spectrum
J  E -3.33+/-0.01
J  E -2.81+/-0.02
Atmospheric Database
Atmospheric data
of the selected
nights in this
analysis:
<HAL> = 27 km
<VAOD> = 0.035
Acceptances & Aperture
Rmc(Ei) / Gmc(Ei)
Acceptances from simulations
broken up into 3 datasets.
A** Rmc(Ei) / Gmc(Ei)
Average instant aperture
(in km2 sr) for all 3 datasets.
Exposure
A* * t * Rmc(Ei) / Gmc(Ei)
Exposure (in 104 km2 sr s) with fit.
A* * t * Rmc(Ei) / Gmc(Ei)
‘Smoothed’ exposure
(in 104 km2 sr s).
• We observe the ‘ankle’ in the HiRes-2 spectrum
at 1018.5 eV.
• The HiRes-2 result is in close agreement with
HiRes-1 and Fly’s Eye.
• The HiRes-2 spectrum is consistent with the
‘second knee’ and GZK flux suppression.
• The combined monocular HiRes spectra show evidence
for a break above 1019.8 eV. The Poisson probability for
continuation of the spectrum with unchanged slope
from the HiRes monocular data is 2.4 * 10-4 .
HiRes-2 Composition Measurement
HiRes/MIA & HR stereo Composition.
HiRes-2 Composition
We can extend composition
analysis down to about
1017.5 eV with HiRes-2
data.
HiRes vs. Auger FD
• 2 eyes, 22 / 42 spherical
mirrors
• azimuth ~360, elevation
3 - 17 / 3-31
• 2 eyes (so far), 6
spherical mirrors each
• azim. 180, el. 28.6
• Schmidt optics
•
•
•
•
•
•
•
•
mirror radius 1.3 m
16x16 PMT per mir.
Pixel size: 1 x 1
UV filter
• Sample&Hold / FADC @
10 MHz
mirror radius 3.4 m
20 x 22 PMT per mir.
pixel size: 1.5 x 1.5
UV filter, Winston cones
• FADC @ 10 MHz
Phototube Calibration
• Absolute calibration using a
portable light-source (“RXF”),
that is carried to both sites.
– calibration of RXF in the lab
using HPDs.
=> +/- 10% uncertainty in energy
scale.
pe = qe * ce * A * 
 = G * pe
 = G * √(*pe)
pe =  * (/) 2
• Relative calibration at the
beginning and end of each
nightly run.
– using YAG laser
– optical fibers distribute the
laser signal to all mirrors.