Precision measurement of the (e$^+$ + e$^-$) flux in primary cosmic rays from 0.5 GeV to 1 TeV with the Alpha Magnetic Spectrometer on the International Space Station

0.65 MB·

by  Manuela Vecchi

XXV European Cosmic Ray Symposium, Turin, Sept. 4-9 2016 1
Precision measurement of the (e+ + e−) flux in primary cosmic rays from 0.5
GeV to 1 TeV with the Alpha Magnetic Spectrometer on the International Space
Station
M. Vecchi on behalf of the AMS-02 Collaboration
Instituto de F´ısica de Sa˜o Carlos (IFSC), Universidade de Sa˜o Paulo, CP 369, 13560-970, Sa˜o Carlos, SP, Brazil
We present a precise measurement of the combined electron plus positron flux from 0.5 GeV to
1 TeV, based on the analysis of the data collected by the Alpha Magnetic Spectrometer during the
first 30 months of operations aboard the International Space Station. The statistics and the high
resolution of AMS-02 detector provide a precise measurement of the flux. The flux is smooth and
reveals new and distinct information. Above 30.2 GeV, the combined electron plus positron flux
can be described accurately by a single power law.
7
1
0 I. INTRODUCTION AND DETECTOR A. Lepton-hadron separation
2
LAYOUT

n
Electronsandpositronsonlyaccountforatinyfrac-
a
tion of the cosmic rays: e− are ∼10−2 less abundant
J
thanprotons,whilee+ are∼10−4 lessabundantthan
9 TheAlphaMagneticSpectrometerisageneralpur-
protons. However, the measurement of their fluxes

] poseparticlephysicsdetector,operatinginspacesince canprovideimportantinformationstounderstandthe
E May 2011. It will achieve a unique long duration nearby universe, as their detection horizon is limited
H mission, aiming at performing antimatter and dark to few kiloparsecs, due to energy losses. Three main
matter searches, as well as cosmic ray composition
. sub-detectors provide clean and redundant identifica-
h and flux measurements [1]. The experiment is in-
tion of positrons and electrons with independent sup-
p stalledonboardtheInternationalSpaceStation(ISS),
pression of the proton background. These are the
-
o that follows a Low Earth Orbit at about 400 km al- TRD(abovethemagnet),theECAL(belowthemag-
r titude with respect to the Earth surface, well located
net) and the tracker. The matching of the ECAL en-
t
s to detect cosmic particles before they interact with ergy and the momentum measured with the tracker
a the outer layers of the atmosphere. This makes the
[ (E/p inthefollowing)greatlyimprovestheprotonre-
ISSonethemostinterestingenvironmentstoperform jection. To differentiate between e± and protons in

1 cosmic rays studies. The measurements presented in
theTRD,anestimatorformedbytheratioofthelog-
v this document are based on the data collected dur-
likelihood probability of the e± hypothesis to that of
2 ing the first 30 months of operations of the detec-
the proton hypothesis in each layer is used. The pro-
1
tor, from May 19th 2011 to November 26th 2013. In
2 ton rejection power of the TRD estimator at 90% e±
this period 41×109 cosmic ray events were detected.
2 efficiency is 103 to 104 [8], as estimated using flight
The detector is composed of several sub-detectors, as
0 data. To cleanly identify electrons and positrons in
1. showninfigure1. Thesilicontracker[2]measuresthe theECAL,aBoostedDecisionTreeestimatorisbuilt
trajectory and absolute charge |Z| of cosmic rays by
0 using the 3D shower shape. The ECAL proton rejec-
performing multiple measurements of the coordinates
7 tion power reaches 104 when combined with the E/p
1 and energy loss. Together with the 0.14 T perma- matching requirement.
: nentmagnet,thetrackermeasurestheparticlerigidity
v
R=pc/Ze,wherepisthemomentum. TheTransition
i
X RadiationDetector(TRD)[3]identifiestheparticleas
r an electron/positron. The four layers of the Time of II. THE COMBINED (e+ + e−) FLUX
a Flight(TOF)[4]measuretheparticleschargeanden- MEASUREMENT
sure that the particle is downward-going. The high
efficiency (∼ 99.999%) anti-coincidence counters [5] The data collected during the first 30 months of
inside the magnet bore are used to reject particles operations of the detector were analysed to provide
outside the geometric acceptance. The Ring Imaging precise measurements of the positron fraction [9] and
CHerenkov detector (RICH) [6] measures the parti- the individual positron and electron fluxes [10]. The
cles charge and velocity. The imaging Electromag- positron flux have been measured up to 500 GeV
neticCalorimeter(ECAL)[7]identifiestheparticleas and of the electron flux up to 700 GeV. These re-
an electron/positron and measures its energy. sults generated widespread interest and discussions
The AMS-02 detector has been extensively calibrated on the origin of high-energy positrons and elec-
using a test beam at CERN with e− and e+ from 10 trons [11] [12] [13].
to 290 GeV, with protons at 180 and 400 GeV, and In this document, based on the published result [14],
with π± from 10 to 180 GeV. we present a dedicated measurement of the combined
eConf TBA
2 XXV European Cosmic Ray Symposium, Turin, Sept. 4-9 2016
1
TRD
TOF
2
T 3-4
E
Tracker
N C
G C 5-6
A A
M 7-8
TOF
RRIICCHH
z
9
y ECAL
FIG. 1: A 369 GeV positron event as measured by the AMS detector on the ISS, in the (y-z) plane.
(e+ + e−) up to 1 TeV. atmosphere, the minimum energy within the bin is
The isotropic flux of cosmic rays electrons and required to exceed 1.2 times the geomagnetic cutoff.
positrons in each energy bin E, of width ∆E, is given Over a sample of well reconstructed particles with a
by [14]: singletrackinthetrackerpassingthroughtheECAL,
theidentificationofsignaleventsisperformedbyfirst
N (E) applyingafixedcutintheECALestimatortofurther
Φ(E)= e (1)
A ·T(E)·∆(E) reduce the proton background. The number of signal
eff
and background event is estimated for each energy
whereN (E)isthenumberofelectronspluspositrons bin performing a template fit procedure, described in
e
with energy between E and E +∆E, A is the ef- [14]. In total, 10.6 ×106 events are identified as elec-
eff
fective acceptance, T(E) is the exposure time. The trons and positrons with energies from 0.5 GeV to
flux is measured in 74 energy bins from 0.5 to 1 TeV, 1 TeV. A major experimental advantage of the com-
and the bin width is chosen to be at least two times bined flux analysis compared to the measurement of
the energy resolution. The bin-to-bin migration error the individual positron and electron fluxes, especially
is 1% at 1 GeV decreasing to 0.2% above 10 GeV. athighenergies,isthattheselectiondoesnotdepend
The effective acceptance A is the product of the onthesignofthecharge,implyinghigherselectionef-
eff
detector geometric acceptance (∼ 500 cm2sr) and the ficiency. Consequently, this measurement is extended
selection efficiency, estimated with simulated events to 1 TeV with less overall uncertainty over the entire
and validated with a pure sample of electron events energy range.
identifiedinthedata. Theexposuretimeisevaluated The absolute energy scale is verified using minimum
as a function of energy and it takes into account the ionizing particles and the ratio between the energy,
lifetime of the experiment which depends on its or- measured by the ECAL, and the momentum, mea-
bit location and on the geomagnetic cutoff [15]. To suredbythetracker. Theseresultsarecomparedwith
identify downward-going particles of charge one, cuts thetestbeamvalueswheretheenergybeamisknown
areappliedonthevelocitymeasuredbytheTOFand to high precision. Between 10 and 290 GeV (Test
on the charge reconstructed by the tracker and the Beam energies), the uncertainty on the absolute scale
TRD. The energy deposited in the ECAL is also used is ∼ 2%, while below 10 GeV it increases to 5 % at
to reject events compatible with a minimum ionising 0.5 GeV and above 290 GeV to 5 % at 1 TeV. This
particle. To reject positrons and electrons produced is treated as an uncertainty on the bin boundaries.
by the interaction of primary cosmic rays with the The statistical error dominates above 140 GeV, as it
eConf TBA
XXV European Cosmic Ray Symposium, Turin, Sept. 4-9 2016 3
FIG. 2: AMS-02 combined positrons plus electrons flux, rescaled by the cube of energy, as a function of energy [14],
together with the most recent measurements from other experiments. See [14] for the complete list of references
is shown in the table 1 in [14]. is expressed in GeV and C is the normalisation con-
Figure 2 shows the AMS-02 combined electrons plus stant. The resulting energy dependence of the fit-
positrons flux, rescaled by the cube of the energy, as ted spectral index is shown in figure 3, where the
a function of energy, together with previous measure- shading indicates the correlation between the neigh-
ments. Given the high statistics and high precision of bouring points due to sliding energy window. In-
the measurement, the spectral index of the combined terestingly, the combined (e+ +e−) flux can be de-
flux has also been measured and, for energies higher scribed by a single power law above 30.2 GeV, with
than 30 GeV, it is found to be compatible with a sin- γ =−3.170±0.008(stat+syst)±0.008(energy scale).
gle power law. The flux, measured from 0.5 to 1 TeV, is smooth and
The measurement of the separate fluxes [10] shows reveals new and distinct information. No structures
thatelectronandpositronfluxesaredifferentinmag- were observed.
nitude and in their energy dependence. Above 20 Acknowledgments
GeV,thepositronfluxissignificantlyharderthanthe
electron flux, implying that the observed rise in the
The author is grateful to the S˜ao Paulo Research
positron fraction is due to an excess of positrons and
Foundation(FAPESP)forthefinancialsupport(grant
not to a loss of electrons. This indicates that high
n. 2014/19149-7 and n. 2014/50747-8). The author
energy positrons have a different origin from that of
is thankful to the organisers of the ECRS conference
electrons. To quantitatively study the energy depen-
for their kind availability.
dence of the flux in a model independent way, the
flux is fit with a spectral index γ as: Φ(e+ +e−) =
CEγ over a sliding energy window. The energy E
[1] S. Della Torre, these proceedings (2010) 237-249
(arXiv:1612.08441v1) [7] C.Adloffetal,Nucl.Inst.Meth.A714(2013)147-154
[2] B. Alpat et al., Nucl. Instr. Meth. A 613 (2009) 207 [8] L. Accardo et al, Phys. Rev. Lett. 110 141102 (2013)
[3] T. Kirn et al, Nucl. Instr. Meth. A 706 (2013) 43-47 [9] M. Aguilar et al, Phys. Rev. Lett. 113 121101 (2014)
[4] V. Bindi et al, Nucl. Instr. Meth. A 743 (2014) 22-29 [10] M. Aguilar et al, Phys. Rev. Lett. 113 121102 (2014)
[5] Ph. von Doetinchem et al, Nucl. Phys. Proc. Suppl. [11] M. Di Mauro et al, JCAP 04 (2014) 006
197 (2009) 15-18 [12] M. Boudaud et al, Astr. Astroph. 575, A67 (2015)
[6] M. Aguilar-Benitez et al, Nucl. Instr. Meth. A 614 [13] Q. Yuan and X.B. Bi, JCAP 1503 (2015) 033
eConf TBA
4 XXV European Cosmic Ray Symposium, Turin, Sept. 4-9 2016
FIG. 3: (a) the spectral index of the Φ(e+ +e−) flux as a function of energy. The shaded regions indicate the 68%
C.L. intervals including the correlation between neighboring points due to the sliding energy window. (b) Φ(e++e−)
multiplied by the cube of the energy as function of energy, above 30.2 GeV, together with the fit to a single power law.
[14] M. Aguilar et al, Phys. Rev. Lett. 113 221102 (2014) sity Press, London,).
[15] C. Stormer. 1950. The Polar Aurora (Oxford Univer-
eConf TBA