Observation of the Galactic Cosmic Ray Moon shadowing effect with the ARGO-YBJ experiment

0.17 MB·English

by  R. Iuppa

PROCEEDINGS OF THE 31st ICRC, ŁO´DZ´ 2009 1
Observation of the Galactic Cosmic Ray Moon shadowing effect
with the ARGO-YBJ experiment
Roberto Iuppa∗ †, Daniele Martello‡§, Bo Wang¶ and Giovanni Zizzi‡§
on behalf of the ARGO-YBJ Collaboration
∗ INFN Sezione Roma Tor Vergata, via della Ricerca Scientifica 1, Rome - Italy
† Dipartimento di Fisica, Universita´ Roma Tor Vergata, via della Ricerca Scientifica 1, Rome - Italy
‡ INFN Sezione di Lecce, via per Arnesano, 73100 Lecce - Italy
§ Dipartimento di Fisica dell’Universita` del Salento, via per Arnesano, 73100 Lecce - Italy
0 ¶ Key Laboratory of Particle Astrophysics, IHEP - Chinese Academy of Science, P.O. Box 918, 100049 Beijing, P.R. China
1
0
2

n Abstract. Cosmic rays are hampered by the Moon resolution and its position allows the evaluation of the
a andadeficitinitsdirectionisexpected(theso-called absolute pointing accuracy of the detector. In addition,
J
Moon shadow). The Moon shadow is an important positively charged particles are eastward deflected, due
4
method to determine the performance of an air to the geomagnetic field, with an energy dependence

shower array. In fact, the westward displacement of ∆θ ∼1.6◦Z/E . As a consequence, the observation
] TeV
E the shadow center, due to the propagation of cosmic ofthedisplacementoftheMoonprovidesadirectcheck
H rays in the geomagnetic field, allows to calibrate the of the relation between shower size and primary energy.
. energy scale of the primary particles observed by Therefore, the analysis of the Moon shadow allows the
h
the detector. In addition, the shape of the shadow calibration of the performance of an EAS array.
p
- allows a measurement of the angular resolution and The same shadowing effect can be observed in the
o the position of the deficit at high energy allows the direction of the Sun but the interpretation of the shad-
r
t evaluation of the pointing accuracy of the detector. owing phenomenology is more complex. In fact, the
s
In this paper we present the observation of the displacement of the shadow from the apparent position
a
[ galactic cosmic rays Moon shadowing effect per- of the Sun could be explained by the joint effects of

formed by the ARGO-YBJ experiment in the multi- the geomagnetic field and of the solar and Interplane-
1
v TeV energy region. The measured angular resolution tary Magnetic Fields, whose configuration considerably
3 asafunctionoftheshowersizeiscomparedwiththe changes with the phases of the solar activity cycle [3].
3 expectations from a MC simulation. Results about the Sun shadow observation with the
5 Keywords: Moon Shadow observation, Cosmic ARGO-YBJ experiment are discussed in [4].
0
Rays, ARGO-YBJ experiment Inthispaperwepresenttheobservationofthegalactic
.
1 cosmic rays Moon shadowing effect carried out by
0 I. INTRODUCTION the ARGO-YBJ experiment. We report on the angular
0
TheangularresolutionisacriticalfeatureofanExten- resolutionofthedetectorinthemulti-TeVenergyregion.
1
: sive Air Shower (EAS) array in gamma-ray astronomy. The pointing error is also investigated.
v
In fact, the rejection of the nearly isotropic background
Xi of charged cosmic rays is mainly performed by im- II. THEARGO-YBJEXPERIMENT
r proving the angular resolution, thus reducing the source The ARGO-YBJ detector, located at the YangBaJing
a
region extension. Hence the tuning of a firm calibration Cosmic Ray Laboratory (Tibet, P.R. China, 4300 m
technique of the angular resolution is mandatory. a.s.l.),istheonlyexperimentexploitingthefullcoverage
The CYGNUS experiment in 1991 reported the first approach at very high altitude. The detector is consti-
determination of the angular resolution of an EAS de- tuted by a central carpet ∼74×78 m2, made of a single
tector by exploiting the analysis of the shadow of the layerofResistivePlateChambers(RPCs)with∼92%of
Moon [1]. In fact, as pointed out in 1957 by Clark [2], active area, enclosed by a partially instrumented guard
thecosmicraysarehamperedintheirpropagationtothe ring that extends the detector surface up to ∼100×110
EarthduetotheMoon’spresenceandadeficitofevents m2. The apparatus has a modular structure, the basic
in its direction is expected: the so-called Moon shadow. dataacquisitionelementbeingacluster(5.72×7.64m2),
At high energy, the Moon shadow would be observed namely a group of 12 RPCs (2.80×1.25 m2 each).
by an ideal detector as a 0.26◦ wide circular deficit of Each chamber is read by 80 strips of 7×62 cm2 (the
events, centered on the Moon position. The deviation spatial pixel), logically organized in 10 independent
from such an ideal case gives us information about the pads of 56×62 cm2 representing the time pixel of the
Point Spread Function (PSF) of the detector. The shape detector. The RPCs are operated in streamer mode with
of the deficit allows the measurement of the angular a standard gas mixture (Argon 15%, Isobutane 10%,
2 R. IUPPA et al. MOON SHADOW OBSERVATION
TetraFluoroEthane75%),theHighVoltagesettledat7.2
kV ensures an overall efficiency of about 96% [5]. The h 4
central carpet contains 130 clusters (hereafter ARGO- Nort 3 0
1firfiswNnoeh3riocreo0AtTtvohdhw)aherleldeetmpatherecoanebvdedAtdcaeen.owRlrnattrTrrhthaGeatsehecohlOagettlecihefniv-uvaesdYfecrlipnulnpeBtaslgehnuldutJtietseraardewftaleaeaendlxirtcccuterpctethomicooeaviotornrraborripfledemriaescri∼dtneiotoiscina6irnmfonett7siemcfinet0Jsrtsr0iuuptweosaancodtnmrinteasntpdeebt2od2hdadl.foe0detwho0tsrphedef6Noeoac.1tstfopipaSi5mart4rid3idin2toeimacn0≥nckeogalinourfnN2ysfsgd0.ttataaer0nhritra7ygaeest uth ---023211 -------3322115505050
o -40
the multiplicity trigger threshold Ntrig ≥20 and a duty S-4
cycle ∼90%: the trigger rate is about 3.6 kHz. -4 -3West - 2 - 1 0 1 2 E3ast 4
The reconstruction of shower parameters is split into
the following steps. First the shower core position is Fig. 1: Moon shadow significance map observed by the
derived with the Maximum Likelihood method from the ARGO-YBJdetectorin2063hourson-sourceforevents
lateral density distribution of the secondary particles. with Nstrip ≥30 and zenith angle θ < 50◦. The color
In the second step, given the shower core position, the scale gives the statistical significance.
shower axis is reconstructed by means of an iterative
weighted planar fit being able to reject the time values
belonging to the non-gaussian tails of the arrival time ascension, keeping unchanged the declination. A new
distributions. A conical correction with a slope fixed to sky map (background map) is built by using 10 such
α = 0.03 rad is applied to the surviving hits in order to fake events for each real one, so that the statistical error
improve the angular resolution [6]. on the background can be kept small enough.
With the equi-zenith angle method 6 off-source bins
III. MONTECARLOSIMULATION aresymmetricallyalignedonbothsidesoftheon-source
The air showers development in the atmosphere has field, at the same zenith angle. The off-source bins are
been generated with the CORSIKA v. 6.500 code in- each set at an azimuth distance 5◦/sinθ from the on-
cluding the QGSJET-II.03 hadronic interaction model source bin, where θ is the zenith angle of the Moon
forprimaryenergyabove80GeVandtheFLUKAcode position.Other off-source bins arelocated every 5◦/sinθ
for lower energies [7]. Cosmic ray spectra of p, He and from the nearest off-source bins. The average of the
CNO have been simulated in the energy range from event densities inside these bins was taken to be the
30 GeV to 1 PeV following [8]. The relative fractions background.
(in % of the total) after triggering by the ARGO- Tomaximizethesignaltonoiseratio,thebinsarethen
YBJ detector for events with N ≥30 are: p∼88%, groupedoveracircularareaofradiusψ,i.e.everybinis
strip
He∼10%, CNO∼2%. About 3·1011 showers have been filledwiththecontentofallthesurroundingbinswhose
distributedinthezenithangleinterval0-60degrees.The center is closer than ψ from its center. The value of ψ
secondaryparticleshavebeenpropagateddowntoacut- is related to the angular resolution of the detector, and
off energy of 1 MeV. The experimental conditions have corresponds to the radius of the observational window
beenreproducedviaaGEANT3-basedcode.Theshower that maximizes the signal to noise ratio, which in turn
core positions have been randomly distributed sampling depends on the number of fired pads of the event: when
in energy-dependent area up to 103×103 m2, centered the PSF is a Gaussian with rms σ, ψ = σ ·1.58 and
on the detector. contains ∼72% of the events. Finally, the integrated
background map is subtracted from the corresponding
IV. DATAANALYSIS integrated event map, thus obtaining the ”source map”.
For the analysis of the shadowing effect a 10◦× 10◦ Foreachbinofsuchamap,thedeficitsignificancewith
sky map in celestial coordinates (right ascension and respect to the background is calculated according to the
declination) with 0.1◦×0.1◦ bin size, centered on the Li and Ma formula [12]. Notice that in the integrated
Moon location, is filled with the detected events. The maps neighboring bins are correlated.
background is evaluated with both the time swapping The analysis reported in this paper refers to events
[10] and the equi-zenith angle [11] methods. collected after the following event selection: (1) each
With the time swapping method, N ”fake” events event should fire more than 30 strips on the ARGO-
are generated for each detected one, by replacing the 130 central carpet to avoid any threshold effect; (2) the
measured arrival time with new ones. These events zenith angle of the shower arrival direction should be
are randomly selected within a 3 hours wide buffer of less than 50◦; (3) the reconstructed core position should
recorded data. Swapping the time is swapping the right be inside an area 250×250 m2 centered on the detector;
PROCEEDINGS OF THE 31st ICRC, ŁO´DZ´ 2009 3
deficit count-220000000 30-60 -deficit count20000 100-300 North 34 0
deficit count--11-----201218640000000000000000000000000-6 -5 -4 -360-2-10-10 0(α - 1αm) c2os(δ3m) (°4) deficit count---864-5500000000000000-3 -2 -1300-0500(α1 - αm) co2s(δm) (°)3 h --20211 ---11550
-3000 -1000 out-3
-4000 -1500 S -20
-5000
-4
-6000 -2000 -4 -3 -2 -1 0 1 2 3 4
-6 -5 -4 -3 -2 -1 0(α - 1αm) c2os(δ3m) (°4) -3 -2 -1 0 (α1 - αm) co2s(δm) (°)3 West East
Fig. 2: Deficit counts measured around the Moon pro- Fig. 3: Moon shadow significance map observed by the
jectedalongtheEast-Westaxisfordifferentmultiplicity ARGO-YBJdetectorin2063hourson-sourceforevents
bins. with N ≥1000 and zenith angle θ <50◦. The color
strip
scale gives the statistical significance.
(4) the reduced χ2 of the final temporal fit should be
less than 100 ns2. According to our simulation studies,
E50 ∼ 30 TeV) Moon shadow position is centered in
the median energy of the selected protons firing 30 ÷ p
the East-West direction but we observe a residual shift
60 strips is E50 ≈1.4 TeV (mode energy ∼0.30 TeV).
towards the North. Since the displacement along the
In Fig.1 the Moon shadow observed with all data
North-South axis is not affected by the geomagnetic
recorded since June 2006 (2063 hours on-source) for
field at the Yangbajing latitude [9], we are able to
events with N ≥30 and zenith angle θ < 50◦ is
strip investigate this pointing error without the Moon shadow
shown. The statistical significance of the observation is
simulationasafunctionofthemultiplicity.Theanalysis
about 43 standard deviations.
has been performed both with the time-swapping and
V. RESULTS the equi-zenith angle methods: the results are in good
The deficit counts observed around the Moon pro- agreementandsuggestthatthereisaresidualsystematic
jected to the East-West axis are shown in Fig. 2 for 4 shift towards North of (0.20±0.05)◦, independent of
multiplicity bins. We use the events contained in an an- multiplicity. As a conservative estimate we assume our
gularsliceparalleltotheEast-Westaxisandcenteredto systematic errors to coincide with this displacement
the observed Moon position. The widths of these bands in both North-.South and East-West projection. These
are function of the N -dependent angular resolution: upper limits for the systematic errors are however much
strip
±3.3◦ in30≤N <60,±2.6◦ in60≤N <100, smaller than our angular resolution, at least for the bulk
strip strip
±2.0◦ in 100≤N < 300, ±1.5◦ in 300≤N < of data, and can therefore be neglected in the point
strip strip
500. As an expected effect of the geomagnetic effect, source searches.
the profile of the shadow is broadened and the peak IntheFig.4thedisplacementoftheMoonshadowin
positions shifted westward as the multiplicity (i.e., the theEast-Westdirectionisshown.Twodifferentmethods
cosmic ray primary energy) decreases. We note that in agreewitheachotherquitewell.Acomparisonbetween
thelowestmultiplicitybin(N =30-60)theMoonis the measurement and the simulations allows to attribute
strip
shiftedbyabout1◦,asexpectedforaprimaryofrigidity this displacement to an absolute energy calibration of
1.6TeV/Z:thisisthefirsttimethatanEAS-arrayisable the detector.
to detect showers with such a low primary energy. The PSF of the detector, studied in the North-South
The best procedure to evaluate the pointing accuracy projection, not affected by the geomagnetic field, is
is to observe the position of the Moon shadow pro- GaussianforN ≥100,whileforlowermultiplicities
strip
duced by high-energy cosmic rays which are negligibly it can be described with an additional Gaussian, which
affectedbythegeomagneticfield.Forprotonsof30TeV contributes for about 20%. For these events the angular
weexpectadeflectionofabout0.05◦.Forheaviernuclei resolutioniscalculatedastheweightedsumoftheσ2 of
this deflection will increase but as the composition of eachgaussian.InFig.5themeasuredangularresolution
cosmic rays in this energy range is dominated by the is compared to expectations from a MC simulation as
light component (nuclei heavier than CNO contribute to a function of the multiplicity. As can be seen, the
theratelessthan3%inthewholestripmultiplicityrange values are in fair agreement: the angular resolution σ
[13]) we expect only a small contribution from heavy of the ARGO-YBJ experiment for cosmic ray-induced
ionstotheblurringoftheMoonshadow.Ascanbeseen airshowersislessthan0.6◦ forN ≥300.Theeffect
strip
fromtheFig.3,theobservedhighenergy(N ≥1000, of the finite angular width of the Moon on the angular
pad
4 R. IUPPA et al. MOON SHADOW OBSERVATION
EE [[TTeeVV]]//ZZ
5500 multi-TeV energy region with high statistical signifi-
11 1100 110022
°α) (0.5 ctiamnec-es.wTahpepianngamlyestihsohdaasnbdetehnepeqerufio-zrmeneidthbaontghlewmitehthtohde
∆ in order to investigate possible biases in the background
0
calculation. The measured angular resolution is in good
agreement with MC simulations, making us confident
-0.5 in the reconstruction algorithms, we can further find an
absolute energy calibration of the detector.
-1 REFERENCES
EW displacement
[1] D.E.Alexandreasetal.,Phys.Rev.D431735,1991.
equi-zenith
-1.5 [2] G.W.Clark,Phys.Rev.108450,1957.
time-swapping [3] M.Amenomorietal.,ApJ 5411051,2000.
[4] F.R.Zhuetal.,theseproceedings
[5] G.Aiellietal.,NIM A56292,2006.
102 103 104
number of fired strips [6] G. Di Sciascio et al., Proc. of 30th ICRC 2009, (preprint:
arXiv:0710.1945).
Fig. 4: Observed displacement of the Moon shadow in [7] D.Hecketal.,ReportFZKA60191998.
the East-West direction as a function of multiplicity. [8] B.Wiebel-Soothetal.,A&A330389,1998.
[9] G.DiSciascioandR.Iuppa,theseproceedings
The upper scale refers to the median energy of rigidity
[10] D.E.Alexandreasetal.,NIM A311350,1992.
(TeV/Z) in each multiplicity bin (shown by the horizon- [11] M.Amenomorietal.,Phys.Rev.D472675,1993.
tal errors). The experimental data calculated with two [12] T.LiandY.Ma,ApJ,272317,1993.
[13] S.Catalanottieral.,theseproceedings.
different methods (see text).
E [TeV]/Z
1 10 50
g.)2.2
e Angular resolution
d 2
σ( MC expectation
1.8
Data
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
102 103
number of fired strips
Fig. 5: Measured angular resolution of the ARGO-YBJ
detector compared to expectations from MC simulation
as a function of the multiplicity. The upper scale refers
to the median energy of rigidity (TeV/Z) in each multi-
plicity bin (shown by the horizontal errors).
resolution (less than 5% if σ >0.4◦ and only 1.7% if
σ >0.7◦) is discussed in [9]. From MC simulations we
draw that the angular resolution for γ-induced showers
at the threshold is at least 30% lower due to their better
definedtimeprofile[6].Anewreconstructionalgorithm
basedonaweightedfitofthetimeprofileinunderstudy
in order to improve the angular resolution.
A new reconstruction algorithm based on a weighted
fit of the time profile is under study in order to improve
the angular resolution.
VI. CONCLUSIONS
The galactic cosmic ray Moon shadowing effect has
been observed by the ARGO-YBJ experiment in the