Results from the ULTRA experiment in the framework of the EUSO project

Abstract

The detection of Cerenkov light from EAS in a delayed coincidence with fluorescence light gives a strong signature to discriminate protons and neutrinos in cosmic rays. For this purpose, the ULTRA experiment has been designed with 2 detectors: a small EAS array (ETscope) and an UV optical device including wide field (Belenos) and narrow field (UVscope) Cerenkov light detectors. The array measures the shower size and the arrival direction of the incoming EAS, while the UV devices, pointing both to zenith and nadir, are used to determine the amount of direct and diffused coincident Cerenkov light. This information, provided for different diffusing surfaces, will be used to verify the possibility of detecting from Space the Cerenkov light produced by UHECRs with the EUSO experiment, on board the ISS.

Full text

Page 1

29th International Cosmic Ray Conference Pune (2005) 00, 101–104
Results from the ULTRA experiment in the framework of the EUSO
project
G. Agnetta
J. Chauvin
S. Giarrusso
A. Mangano
J.C. Silva
(a) INAF-IASF, Palermo, Italy
(b) LIP, Lisbon, Portugal
(c) INFN, Catania, Italy
(d) LPSC-Grenoble, France
(e) University of Turin and INFN, Torino, Italy
(f) DIFTER, University of Palermo, Italy
(g) INAF-IFSI and INFN, Torino, Italy
Deceased
Presenter: P. Vallania (Piero.Vallania@to.infn.it), ita-vallania-P-abs1-he12-poster
The detection ofˇCerenkov light from EAS in a delayed coincidence with fluorescence light gives a strong
signature to discriminate protons and neutrinos in cosmic rays. For this purpose, the ULTRA experiment
has been designed with 2 detectors: a small EAS array (ETscope) and an UV optical device including wide
field (Belenos) and narrow field (UVscope)ˇCerenkov light detectors. The array measures the shower size and
the arrival direction of the incoming EAS, while the UV devices, pointing both to zenith and nadir, are used to
determinethe amountofdirect anddiffusedcoincidentˇCerenkovlight. This information,providedfordifferent
diffusing surfaces, will be used to verify the possibility of detecting from Space theˇCerenkov light produced
by UHECRs with the EUSO experiment, on board the ISS.
?, P. Assis
, G. D’Al´ ı Staiti
?, B. Biondo
????, P. Brogueira
?, A. Cappa
?, O. Catalano
????,
?
???????, M. Dattoli
?, M.C. Espirito-Santo
?, L. Fava
?, P. Galeotti
?,
????, G. Gugliotta
???????, G. La Rosa
????, D. Lebrun
?
, M.C. Maccarone
????,
????, L. Melo
?, S. Moreggia
, B. Tom` e
?
, M. Pimenta
?, F. Russo
?, O. Saavedra
?, P. Scarsi
????,
?, P. Stassi
??, P. Vallania
? and C. Vigorito
?(EUSO Collaboration)
?
1. Introduction
The study of the extreme region of the primary cosmic ray spectrum represents one of the most challenging
research in this field. The involved physics items range from astrophysics (acceleration mechanism and propa-
gatingspaceproperties)tonuclearandsub-nuclearphysics(interactionwith theatmospherenucleiin anenergy
and pseudo rapidity region very far from actual and future colliders), bearing in mind that we are dealing with
particles carrying even 60 Joules in one single particle. At these energies, also very exotic processes, like mini
blackholesevaporationorhighmassparticlescreation,asHiggsbosons,couldbethestandardbehavior. Dueto
the very low flux (
1 particle/(km
eV), very large detectors and years of stable operations are required to record a significant amount of data. The
Auger project, actually under constructionin Argentina,has been designed following this traditional approach,
and with 3000 km
dimensions, only very few events per year are expected at E
present, and also the ultra high energy neutrinos, carrying cosmological informations on the early Universe,
cannot be detected with enough statistics. A new detector concept must be exploited, and the observation from
Space of the fluorescent andˇCerenkov signals produced along the shower development and its impact with
the Earth’s surface, as proposed by the EUSO Collaboration [1], seems the right answer: with the proposed
field of view (f.o.v.) of 30
of atmosphere can be continuously monitored during night. The feasibility of the fluorescent light detection
and the background measurements have been studied elsewhere [2]; the goal of the ULTRA experiment (Uv
?? century) beyond the Greisen-Zatsepin-Kuzmin cutoff at E
???????
5x10
???
? detection area probably represents the ultimate detector of this kind. Despite its gigantic
10
???? eV if the GZK absorption mechanism is
?? at the ISS altitude, a surface of
? ?"!$#&% km
? and a target mass of
?'?(!)#"??? tons

Page 2

102
G. Agnetta et al.
Light Transmission and Reflection in the Atmosphere) described in this paper is to verify the possibility of
detecting the reflected/diffusedˇCerenkov light produced by the EAS impacting on different Earth’s surfaces
(mainly ocean water).
2.The experiment
The ULTRA experiment is made by a traditional EAS array, called ETscope, and two different kind of
ˇCerenkov light detectors, called UVscope and Belenos. In order to validate the collected signals, minimiz-
ing the possibility of fake events, both direct and diffusedˇCerenkov light is simultaneously detected, and the
diffusedsignalis seen bothfromclose tothediffusingsurface,witha largefield detector,andfarfromthearray,
to avoid the particle contamination, with a narrow f.o.v. detector. Preliminary measurements have been done
at mountain altitude, in Mont-Cenis, to increase theˇCerenkov and electromagnetic signals; then we moved
to Grenoble (France, 210 m a.s.l.) at the LPSC to perform the detector calibration and optimization. Four
ETscope detectors were arranged at the vertex of a square, 54m side, and a fifth detector in the centre to select
the events with the shower core located inside the square. The threshold of each detector was set to 0.3 parti-
cles; further details on the ETscope detector calibration and performances can be found in [3]. TheˇCerenkov
light detectors were located near the central station: the first pair of Belenos (wide field UV detectors), point-
ing to zenith, was mounted on ground, while the second one and a couple of Uvscope units (narrow filed UV
detectors) were mounted on a scaffold pointing to nadir from a 3.5 m height. The Belenos are equipped with a
20 cm Fresnel borosilicate glass lens, with a second lens 8 cm diameter located in the focus of the first one and
used to enlarge the f.o.v. of the device. A 0.5 inch Photomultiplier Tube (PMT) Hamamatsu R960 is located at
a distance of 4 cm from the second lens; with this configuration a f.o.v. of
collecting area (taking into account the lens and PMT glass transmission coefficient, the optical configuration
and the PMT diameter) increasing from 9 cm
The PMT is equipped with a BG3 filter to broadly select the lower wavelength region of theˇCerenkov spec-
trum and to reduce the background; the convolution of the filter transmittance, quantum efficiency and glass
transmission gives a mean value of 9.4%, with a peak at 330 nm when theˇCerenkov spectrum is included. The
UVscope optics is made by a single Fresnel acrylic lens, 46 cm diameter, with a 3.5-inches PMT UV-extended
in the lens focus. The resulting f.o.v. is4
Due to the high background level and for calibration purposes, a layer of highly diffusing material (Tyvek
1025D, with hemisperical reflectivity
80 % in the 300-400 nm range) is placed in the f.o.v. of the nadir UV
detectors. At that time, a double DAQ system was in operation, working simultaneously under LabView DAQ
software: a CAMAC system, with a 2229 LeCroy TDC (250 ps/channel) and a 2249W Lecroy ADC charge
integrator (300 ns integration time), and a 8 bits, 100 MHz, 6 channels flash ADC, under development at LIP
[4]. For this second DAQ system, the total charge and timing are obtained fitting the 10 ns samples; due to the
possibility ofchangingon an eventbyeventbasis the integrationtime to matchthe effectivepulse duration,this
system is veryeffectiveto evidenceverysmall signals as expectedfromthe UV detectors. A simple calculation
of the ratio between the directˇCerenkovsignal measured by the Belenos-up and the diffusedsignal detected by
the Belenos-down gives a value of 0.39 for
depends on the shower arrival direction, but it is largely independent from the distance between the detector
and the diffusing surface.
?
40
? is obtained, with an effective
? for para-axial rays to 25 cm
? for the maximum off-axis angle.
?? .
*
+ =20
? assuming 100% completely diffusing material. This value
3.The MC simulation
EAS generated by primary protons have been developed in the atmosphere using the CORSIKA code [5] with
the QGSJET hadron interaction model. The showers have been simulated for 5 fixed energies, ranging from
10
to 0 m a.s.l.. In order to reduce the computing time, the NKG option has been used; only a small sub-sample
?-, to 10
?-. eV, and arrival direction
+ = 0
? and 20
? (2000 events per point); the observation level has been set

Page 3

Results from the ULTRA experiment in the framework of the EUSO project
103
of showers has been simulated with the EGS andˇCerenkov options to evaluate the energy thresholds of UV
detectors. The CORSIKA output has then been used to calculate the ETscope effective area and shower size
resolution: the shower cores have been sampled over a square area increasing with the energy and ranging
from 2.9x10
including poissonian and experimental fluctuations, and the trigger conditions have been applied; finally the
simulated events have been analyzed as the experimental ones fitting the core location and particle density
measured by each detector with the usual NKG lateral distribution function. For these studies, only internal
events have been chosen, i.e. events with the maximum number of particles detected by the central station;
from the experimental data they correspond to 20 % of the total triggers. Fig. 1 shows the convolution of the
/ m
? at 10
?0, eV to 4.7x10
, m
? at 10
?-. eV; the response of each detector has been simulated,
E (GeV)
5
10
6
10
7
10
-1
Energy Spectrum (s sr GeV)
⊗
Effective Area
-11
10
-10
10
-9
10
o
=0
=20
θ
θ
o
Figure 1. Effective area for internal events convoluted with
the primary spectrum.
Entries 4224Entries 4224
(Ne)
10
log
44.5 55.5 66.57
1
10
2
10
3
10
Figure 2. Shower size spectrum for internal events.
effective area with the primaryspectrum for zenith angles 0
eV, corresponding to the typical detected energy. The detector distances have been indeed optimized to obtain
this value, which matches the threshold energy of theˇCerenkov light detectors. The shower size and core
location resolutions have been calculated using a trial area of 270
minimize the computing time. At E=10
? and 20
? : the mode of the distribution is
*213?4!)#"?-,
5 270 m
35% and
? and the reciprocity method [6] to
6 x
??% eV we obtain
6 N
7 /N
78*90:?;?7 =
6 y
90:?;47<*
6 m.
4.Results
Experimental data have been collected from January to December 2004; a total of 21539 events during 948 h
of measurements have been recorded. Fig. 2 shows the shower size spectrum for the internal events. The
Belenos PMTs have been calibrated using the single photon spectrum; with this calibration and the geo-
metrical acceptance of the optics, theˇCerenkov photon density of the incoming shower can be measured
with the arrival direction provided by the ETscope array. Since two devices are present for each detector,
their coincidence is required; the trigger is given by the 4-fold coincidence of the external ETscope detec-
tors, and the UV devices are acquired passively. Despite the high background level (
sr) in the BG3 filter window,3-4 times the clear night zenithal background) the required 100 ns coin-
cidence with the central station allows us not to be overwhelmed by random signals. Figure 3 shows the
measured photon density at 20 m from the shower core as a function of the shower size N
cation and shower size are determined by the ETscope array; the measured photon density has been scaled
*
3000 ph /(m
? ns
*
7 . Both core lo-

Page 4

104
G. Agnetta et al.
to the quoted distance fitting the simulatedˇCerenkov l.d.f. . Corrections for atmospheric transmission are
not applied at this level. The dotted line is the expectation from Monte Carlo simulation for proton induced
vertical shower. Even if the fluctuations are big, the agreement between the expected values and the measured
mean values is quite good. The square points refer to the CAMAC DAQ, while the circle ones to the LIP
DAQ; a slight systematic difference seems to be present, probably due to the different way to process data.
The threshold energy can be easily obtained, and corre-
sponds to Log
under this value the directˇCerenkov photons are over-
whelmed by the backgroundlight and cannot be seen. For
diffused light, due to the calculated ratio between up and
down, the threshold energy is Log
to the slope of the primary spectrum, we expect that only
14% of the up events have also a diffused signal. The
open symbols are the events with both direct and diffused
ˇCerenkov signal: they are 5 over a total of 161, but with
different observation times (174.3 h for up and 76.6 h for
down). From these events we obtain a mean reflectivity
of 77
80 %. Using the down/up background ratio (which is ex-
pected to be 1 for 100 % reflectivity) a value of 87
% is also obtained. We have performedthe analysis of the
UVscopesignals but, dueto the largephotocathodesize of
its PMTs and the related position in the array, we cannot
be confident that the signals seen by the UVscope detec-
tor are entirely due to the diffusedˇCerenkov light, being
the contamination of the shower particles not negligible (particle crossing the PMTs). Such a problem is not
present in the Belenos detectors since their photocathodearea is about 50 times smaller than the UVscope one.
(Ne)
10
log
4 4.555.566.57
) @ r=20 m
-2
m
C
γ
(
10
log
4.5
5
5.5
6
6.5
7
(r=20 m)
ρ
Expected
Figure 3. Photon density at 20 m from the shower core
as a function of the shower size (see text).
?-?N
7>=@?BAC (E
=
4x10
?0, eV for protons);
?0?N
7D=
5.3, and, due
?
17 %, to be compared with the nominal value of
?
16
5. Conclusions
The two components of the ULTRA detector (ETscope and Belenos) have been calibrated and tested with real
EAS signals, electromagnetic andˇCerenkov. Due to the very high background level, the energy threshold
for UV light detection is very high, with a very limited number of events. Nevertheless we latched onto the
technique. The experiment has been moved to sea surface in Sicily (Capo Granitola, Italy) in April 2005; the
measurements have been resumed in May 2005. The ETscope setup is very close to the Grenoble one, and
the UVscope is now observing the center of the array far from it, with a very low contamination from shower
particles, and the possibility of their discrimination using the timing informations. The firstˇCerenkov signals
diffused by the sea water have already been observed.
References
[1] EUSO Collaboration, www.euso-mission.org/docs/RedBookEUSO 21apr04.pdf(2004).
[2] T. Mineo et al., Proc. 28th ICRC, Tsukuba (2003), 833.
[3] J. Chauvin et al., Proc. 28th ICRC, Tsukuba (2003), 1053.
[4] P. Assis et al., Proc. 28th ICRC, Tsukuba (2003), 947.
[5] D. Heck et al., Karlsruhe Report FZKA 6019 (1998).
[6] G. Battistoni et al., Astrop. Phys. 7, 101 (1997).