M. Aguilar, L. Ali Cavasonza, G. Ambrosi, L. Arruda, N. Attig, F. Barao, L. Barrin, A. Bartoloni, S. Ba¸se˘gmez-du Pree, J. Bates, R. Battiston, M. Behlmann, B. Beischer, J. Berdugo, B. Bertucci, V. Bindi, W. de Boer, K. Bollweg, B. Borgia, M.J. Boschini, M. Bourquin, E.F. Bueno, J. Burger, W.J. Burger, S. Burmeister, X.D. Cai, M. Capell, J. Casaus, G. Castellini, F. Cervelli, Y.H. Chang, G.M. Chen, H.S. Chen, Y. Chen, L. Cheng, H.Y. Chou, S. Chouridou, V. Choutko, C.H. Chung, C. Clark, G. Coignet, C. Consolandi, A. Contin, C. Corti, Z. Cui, K. Dadzie, Y.M. Dai, C. Delgado, S. Della Torre, M.B. Demirköz, L. Derome, S. Di Falco, V. Di Felice, C. Díaz, F. Dimiccoli, P. von Doetinchem, F. Dong, F. Donnini, M. Duranti, A. Egorov, A. Eline, J. Feng, E. Fiandrini, P. Fisher, V. Formato, C. Freeman, Y. Galaktionov, C. Gámez, R.J. García-López, C. Gargiulo, H. Gast, I. Gebauer, M. Gervasi, F. Giovacchini, D.M. Gómez-Coral, J. Gong, C. Goy, V. Grabski, D. Grandi,
M. Graziani, K.H. Guo, S. Haino, K.C. Han, R.K. Hashmani, Z.H. He, B. Heber, T.H. Hsieh, J.Y. Hu, Z.C. Huang, W. Hungerford, M. Incagli, W.Y. Jang, Yi Jia, H. Jinchi, K. Kanishev, B. Khiali, G.N. Kim,
Th. Kirn, M. Konyushikhin, O. Kounina, A. Kounine, V. Koutsenko, A. Kuhlman, A. Kulemzin, G. La Vacca, E. Laudi, G. Laurenti, I. Lazzizzera, A. Lebedev, H.T. Lee, S.C. Lee, C. Leluc, J.Q. Li, M. Li, Q. Li, S. Li, T.X. Li, Z.H. Li, C. Light, C.H. Lin, T. Lippert, Z. Liu, S.Q. Lu, Y.S. Lu, K. Luebelsmeyer, J.Z. Luo, S.S. Lyu, F. Machate, C. Mañá, J. Marín, J. Marquardt, T. Martin, G. Martínez, N. Masi, D. Maurin, A. Menchaca-Rocha, Q. Meng, D.C. Mo, M. Molero, P. Mott, L. Mussolin, J.Q. Ni, N. Nikonov, F. Nozzoli, A. Oliva, M. Orcinha, M. Palermo, F. Palmonari, M. Paniccia, A. Pashnin, M. Pauluzzi, S. Pensotti, H.D. Phan, V. Plyaskin, M. Pohl, S. Porter, X.M. Qi, X. Qin, Z.Y. Qu, L. Quadrani, P.G. Rancoita, D. Rapin, A. Reina Conde, S. Rosier-Lees, A. Rozhkov, D. Rozza, R. Sagdeev, S. Schael, S.M. Schmidt, A. Schulz von Dratzig, G. Schwering, E.S. Seo, B.S. Shan, J.Y. Shi, T. Siedenburg, C. Solano, J.W. Song, R. Sonnabend, Q. Sun, Z.T. Sun, M. Tacconi, X.W. Tang, Z.C. Tang, J. Tian, Samuel C.C. Ting, S.M. Ting, N. Tomassetti, J. Torsti, C. Tüysüz, T. Urban, I. Usoskin, V. Vagelli, R. Vainio, E. Valente, E. Valtonen, M. Vázquez Acosta, M. Vecchi, M. Velasco, J.P. Vialle, L.Q. Wang, N.H. Wang, Q.L. Wang, S. Wang, X. Wang, Z.X. Wang,
Received date : 5 September 2020 Accepted date : 9 September 2020
Please cite this article as: M. Aguilar, L.A. Cavasonza, G. Ambrosi et al., The Alpha Magnetic Spectrometer (AMS) on the international space station: Part II – Results from the first seven years, Physics Reports(2020), doi:https://doi.org/10.1016/j.physrep.2020.09.003.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
The Alpha Magnetic Spectrometer (AMS)
1
on the International Space Station:
2
Part II – Results from the First Seven Years
3
M. Aguilar,
30L. Ali Cavasonza,
1G. Ambrosi,
36L. Arruda,
28N. Attig,
24F. Barao,
284
L. Barrin,
15A. Bartoloni,
42S. Ba¸segmez-du Pree,
18,aJ. Bates,
21R. Battiston,
39, 405
M. Behlmann,
10B. Beischer,
1J. Berdugo,
30B. Bertucci,
36, 37V. Bindi,
20W. de Boer,
256
K. Bollweg,
21B. Borgia,
42, 43M.J. Boschini,
32M. Bourquin,
16E.F. Bueno,
18J. Burger,
107
W.J. Burger,
39S. Burmeister,
26X.D. Cai,
10M. Capell,
10J. Casaus,
30G. Castellini,
148
F. Cervelli,
38Y.H. Chang,
47, 48G.M. Chen,
6, 7H.S. Chen,
6, 7Y. Chen,
16L. Cheng,
229
H.Y. Chou,
48S. Chouridou,
1V. Choutko,
10C.H. Chung,
1C. Clark,
10, 21G. Coignet,
310
C. Consolandi,
20A. Contin,
8, 9C. Corti,
20Z. Cui,
22, 23K. Dadzie,
10Y.M. Dai,
5C. Delgado,
3011
S. Della Torre,
32M.B. Demirk¨oz,
2L. Derome,
17S. Di Falco,
38V. Di Felice,
44,bC. D´ıaz,
3012
F. Dimiccoli,
39P. von Doetinchem,
20F. Dong,
34F. Donnini,
44,bM. Duranti,
36A. Egorov,
1013
A. Eline,
10J. Feng,
10E. Fiandrini,
36, 37P. Fisher,
10V. Formato,
44,bC. Freeman,
2014
Y. Galaktionov,
10C. G´amez,
30R.J. Garc´ıa-L´opez,
27C. Gargiulo,
15H. Gast,
1I. Gebauer,
2515
M. Gervasi,
32, 33F. Giovacchini,
30D. M. G´omez-Coral,
20J. Gong,
34C. Goy,
3V. Grabski,
3116
D. Grandi,
32, 33M. Graziani,
36, 37K.H. Guo,
19S. Haino,
47K.C. Han,
29R.K. Hashmani,
217
Z.H. He,
19B. Heber,
26T.H. Hsieh,
10J.Y. Hu,
6, 7Z.C. Huang,
19W. Hungerford,
2118
M. Incagli,
38W.Y. Jang,
13Yi Jia,
10H. Jinchi,
29K. Kanishev,
39B. Khiali,
44,bG.N. Kim,
1319
Th. Kirn,
1M. Konyushikhin,
10O. Kounina,
10A. Kounine,
10V. Koutsenko,
10A. Kuhlman,
2020
A. Kulemzin,
10G. La Vacca,
32, 33E. Laudi,
15G. Laurenti,
8I. Lazzizzera,
39, 40A. Lebedev,
1021
H.T. Lee,
46S.C. Lee,
47C. Leluc,
16J.Q. Li,
34M. Li,
1Q. Li,
34S. Li,
1T.X. Li,
19Z.H. Li,
622
C. Light,
20C.H. Lin,
47T. Lippert,
24Z. Liu,
16S.Q. Lu,
19Y.S. Lu,
6K. Luebelsmeyer,
123
J.Z. Luo,
34S.S. Lyu,
19F. Machate,
1C. Ma˜
n´a,
30J. Mar´ın,
30J. Marquardt,
26T. Martin,
10, 2124
G. Mart´ınez,
30N. Masi,
8, 9D. Maurin,
17A. Menchaca-Rocha,
31Q. Meng,
34D.C. Mo,
1925
M. Molero,
30P. Mott,
10, 21L. Mussolin,
36, 37J.Q. Ni,
19N. Nikonov,
1F. Nozzoli,
3926
A. Oliva,
8M. Orcinha,
28M. Palermo,
20F. Palmonari,
8, 9M. Paniccia,
16A. Pashnin,
1027
M. Pauluzzi,
36, 37S. Pensotti,
32, 33H.D. Phan,
10V. Plyaskin,
10M. Pohl,
16S. Porter,
2128
X.M. Qi,
19X. Qin,
10Z.Y. Qu,
47L. Quadrani,
8, 9P.G. Rancoita,
32D. Rapin,
16A. Reina
29
Conde,
27S. Rosier-Lees,
3A. Rozhkov,
10D. Rozza,
32, 33R. Sagdeev,
11S. Schael,
130
S.M. Schmidt,
24A. Schulz von Dratzig,
1G. Schwering,
1E.S. Seo,
12B.S. Shan,
4J.Y. Shi,
3431
T. Siedenburg,
1C. Solano,
10J.W. Song,
23R. Sonnabend,
1Q. Sun,
23Z.T. Sun,
6, 732
M. Tacconi,
32, 33X.W. Tang,
6Z.C. Tang,
6J. Tian,
36, 37Samuel C.C. Ting,
10, 1533
S.M. Ting,
10N. Tomassetti,
36, 37J. Torsti,
49C. T¨
uys¨
uz,
2T. Urban,
10, 21I. Usoskin,
3534
V. Vagelli,
41, 36R. Vainio,
49E. Valente,
42, 43E. Valtonen,
49M. V´azquez Acosta,
2735
M. Vecchi,
18M. Velasco,
30J.P. Vialle,
3L.Q. Wang,
23N.H. Wang,
23Q.L. Wang,
536
S. Wang,
20X. Wang,
10Z.X. Wang,
19J. Wei,
16Z.L. Weng,
10H. Wu,
34R.Q. Xiong,
3437
W. Xu,
22, 23Q. Yan,
10Y. Yang,
45H. Yi,
34Y.J. Yu,
5Z.Q. Yu,
6M. Zannoni,
32, 33C. Zhang,
638
F. Zhang,
6F.Z. Zhang,
6, 7J.H. Zhang,
34Z. Zhang,
10F. Zhao,
6, 7Z.M. Zheng,
439
H.L. Zhuang,
6V. Zhukov,
1A. Zichichi,
8, 9N. Zimmermann,
1and P. Zuccon
39, 403 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57
Jour
nal
Pr
e-pr
oof
2
Department of Physics, Middle East Technical University (METU), 06800 Ankara, Turkey
43
3
Univ. Grenoble Alpes, Univ. Savoie Mont Blanc,
44
CNRS, LAPP-IN2P3, 74000 Annecy, France
45
4
Beihang University (BUAA), Beijing, 100191, China
46
5
Institute of Electrical Engineering (IEE), Chinese
47
Academy of Sciences, Beijing, 100190, China
48
6
Institute of High Energy Physics (IHEP), Chinese
49
Academy of Sciences, Beijing, 100049, China
50
7
University of Chinese Academy of Sciences (UCAS), Beijing, 100049, China
51
8
INFN Sezione di Bologna, 40126 Bologna, Italy
52
9
Universit`a di Bologna, 40126 Bologna, Italy
53
10
Massachusetts Institute of Technology (MIT), Cambridge, Massachusetts 02139, USA
54
11
East–West Center for Space Science, University
55
of Maryland, College Park, Maryland 20742, USA
56
12
IPST, University of Maryland, College Park, Maryland 20742, USA
57
13
CHEP, Kyungpook National University, 41566 Daegu, Korea
58
14
CNR–IROE, 50125 Firenze, Italy
59
15
European Organization for Nuclear Research (CERN), 1211 Geneva 23, Switzerland
60
16
DPNC, Universit´e de Gen`eve, 1211 Gen`eve 4, Switzerland
61
17
Univ. Grenoble Alpes, CNRS, Grenoble INP, LPSC-IN2P3, 38000 Grenoble, France
62
18
Kapteyn Astronomical Institute, University of Groningen,
63
P.O. Box 800, 9700 AV Groningen, The Netherlands
64
19
Sun Yat–Sen University (SYSU), Guangzhou, 510275, China
65
20
Physics and Astronomy Department, University of Hawaii, Honolulu, Hawaii 96822, USA
66
21
National Aeronautics and Space Administration Johnson
67
Space Center (JSC), Houston, Texas 77058, USA
68
22
Shandong Institute of Advanced Technology (SDIAT), Jinan, Shandong, 250100, China
69
23
Shandong University (SDU), Jinan, Shandong, 250100, China
70
24
J¨
ulich Supercomputing Centre and JARA-FAME,
71
Research Centre J¨
ulich, 52425 J¨
ulich, Germany
72
25
Institut f¨
ur Experimentelle Teilchenphysik, Karlsruhe
73
Institute of Technology (KIT), 76131 Karlsruhe, Germany
74
26
Institut f¨
ur Experimentelle und Angewandte Physik,
75
Christian-Alberts-Universit¨at zu Kiel, 24118 Kiel, Germany
76
27
Instituto de Astrof´ısica de Canarias (IAC), 38205 La Laguna, and Departamento
77
de Astrof´ısica, Universidad de La Laguna, 38206 La Laguna, Tenerife, Spain
78
28
Laborat´orio de Instrumenta¸c˜ao e F´ısica Experimental
79
de Part´ıculas (LIP), 1649-003 Lisboa, Portugal
80
29
National Chung–Shan Institute of Science and
81
Technology (NCSIST), Longtan, Tao Yuan, 32546, Taiwan
82
30
Centro de Investigaciones Energ´eticas, Medioambientales
83
y Tecnol´ogicas (CIEMAT), 28040 Madrid, Spain
84
31
Instituto de F´ısica, Universidad Nacional Aut´onoma
85
de M´exico (UNAM), Ciudad de M´exico, 01000 Mexico
86
32
INFN Sezione di Milano–Bicocca, 20126 Milano, Italy
87
33
Universit`a di Milano–Bicocca, 20126 Milano, Italy
88
34
Southeast University (SEU), Nanjing, 210096, China
89 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
Jour
nal
Pr
e-pr
oof
35
Sodankyl¨a Geophysical Observatory and Space Physics and
90
Astromony Research Unit, University of Oulu, 90014 Oulu, Finland
91
36
INFN Sezione di Perugia, 06100 Perugia, Italy
92
37
Universit`a di Perugia, 06100 Perugia, Italy
93
38
INFN Sezione di Pisa, 56100 Pisa, Italy
94
39
INFN TIFPA, 38123 Povo, Trento, Italy
95
40
Universit`a di Trento, 38123 Povo, Trento, Italy
96
41
Agenzia Spaziale Italiana (ASI), 00133 Roma, Italy
97
42
INFN Sezione di Roma 1, 00185 Roma, Italy
98
43
Universit`a di Roma La Sapienza, 00185 Roma, Italy
99
44
INFN Sezione di Roma Tor Vergata, 00133 Roma, Italy
100
45
National Cheng Kung University, Tainan, 70101, Taiwan
101
46
Academia Sinica Grid Center (ASGC), Nankang, Taipei, 11529, Taiwan
102
47
Institute of Physics, Academia Sinica, Nankang, Taipei, 11529, Taiwan
103
48
Physics Department and Center for High Energy and High Field
104
Physics, National Central University (NCU), Tao Yuan, 32054, Taiwan
105
49
Space Research Laboratory, Department of Physics and
106
Astronomy, University of Turku, 20014 Turku, Finland
107
Abstract
108
The Alpha Magnetic Spectrometer (AMS) is a precision particle physics detector on the
Inter-109
national Space Station (ISS) conducting a unique, long-duration mission of fundamental physics
110
research in space. The physics objectives include the precise studies of the origin of dark matter,
111
antimatter, and cosmic rays as well as the exploration of new phenomena. Following a 16-year
112
period of construction and testing, and a precursor flight on the Space Shuttle, AMS was installed
113
on the ISS on May 19, 2011. In this report we present results based on 120 billion charged cosmic
114
ray events up to multi-TeV energies. This includes the fluxes of positrons, electrons, antiprotons,
115
protons, and nuclei. These results provide unexpected information, which cannot be explained
116
by the current theoretical models. The accuracy and characteristics of the data, simultaneously
117
from many different types of cosmic rays, provide unique input to the understanding of origins,
118
acceleration, and propagation of cosmic rays.
119 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57
Jour
nal
Pr
e-pr
oof
CONTENTS
120
Introduction
6
121
I. AMS on the ISS
8
122
A. Permanent Magnet
10
123
B. Tracker
11
124
1. Tracker coordinate measurements
11
125
2. Tracker rigidity measurement
15
126
3. Determination of tracker absolute rigidity scale and alignment
17
127
4. Charge-sign identification
20
128
5. Tracker charge measurement
22
129
C. Transition Radiation Detector (TRD)
25
130
D. Time of Flight Counters (TOF)
28
131
E. Anticoincidence Counters (ACC)
32
132
F. Ring Imaging Cherenkov Counter (RICH)
33
133
G. Electromagnetic Calorimeter (ECAL)
34
134
1. ECAL shower reconstruction
36
135
2. ECAL energy reconstruction
37
136
3. Proton rejection using ECAL
39
137
H. Trigger and Data Acquisition
41
138
II. Origins of Cosmic Positrons
45
139
III. Origins of Cosmic Electrons
61
140
IV. Cosmic Protons
74
141
V. Cosmic Antiprotons
77
142
VI. Properties of Cosmic Elementary Particles
82
143
VII. Nuclear Cross Section Measurements
90
144
VIII. Primary Helium, Carbon, and Oxygen Fluxes
96
145
IX. Proton-to-Helium Flux Ratio
100
146
X. Secondary Lithium, Beryllium, Boron Fluxes
147
and Secondary to Primary Ratios
102
148
XI. Properties of Cosmic Helium Isotopes
111
149
XII. Cosmic Nitrogen Flux
116
150
XIII. Primary Neon, Magnesium, and Silicon Fluxes
124
151
XIV. Strangelets
133
152
XV. Time-Dependent Proton and Helium Fluxes
137
153 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
Jour
nal
Pr
e-pr
oof
XVI. Time-Dependent Electron and Positron Fluxes
143
154XVII. Summary
151
155Acknowledgments
152
156References
153
157 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57Jour
nal
Pr
e-pr
oof
INTRODUCTION
158
The Alpha Magnetic Spectrometer (AMS) is a precision particle physics detector on the
159
International Space Station (ISS) conducting a unique, long-duration mission of fundamental
160
physics research in space. The physics objectives include precise studies of the origins of dark
161
matter, antimatter, and cosmic rays as well as the exploration of new phenomena. Prior
162
to the main mission, a precursor flight of AMS, AMS-01, was flown on the Space Shuttle
163
Discovery in June 1998. AMS-01 results were published in Physics Reports, “The Alpha
164
Magnetic Spectrometer (AMS) on the International Space Station: Part I – results from the
165
test flight on the space shuttle” [
1
]. With the experience accumulated from AMS-01, a new
166
state of the art detector, known as AMS-02 or AMS on the ISS, was built and launched on
167
the Space Shuttle Endeavour and installed on the ISS on May 19, 2011 (see Figure
1
). The
168
ISS is orbiting the Earth at an altitude of
∼ 410 km with an orbit inclination of 52 degrees.
169
AMS
FIG. 1. AMS is a unique precision magnetic spectrometer on the ISS. AMS will operate on the ISS for the Station’s lifetime. It is mounted on the ISS with a 12 degree angle to the zenith to prevent that the rotating ISS solar arrays are in the AMS field of view.
There are two kinds of cosmic rays in space:
170
1. Neutral cosmic rays (light rays and neutrinos) that have been studied by many
satel-171
lite (COBE [
2
], EGRET [
3
], WMAP [
4
], Planck [
5
], ROSAT [
6
], Fermi [
7
],
AG-172
ILE [
8
], Chandra [
9
], INTEGRAL [
10
], the Hubble Space Telescope [
11
] and the coming
173 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
Jour
nal
Pr
e-pr
oof
James Webb Telescope [
12
], . . . ) and ground based (ARGO-YBJ [
13
], HAWC [
14
],
174
H.E.S.S. [
15
], MAGIC [
16
], IceCube [
17
], Tibet ASgamma [
18
], LHAASO [
19
], . . . )
175
experiments. Studies of neutral cosmic rays have provided fundamental information
176
about the universe;
177
2. Charged cosmic rays are absorbed in the Earth’s atmosphere and therefore, in the
178
energy range up to multi-TeV, their properties such as charge magnitude and sign,
179
mass, and rigidity can only be studied in space. In the past there have been many
180
excellent experiments with balloons (ATIC [
20
], BESS [
21
], CAPRICE/WiZard [
22
],
181
CREAM [
23
], . . . ) and with satellites (CRIS [
24
], HEAO [
25
], PAMELA [
26
], . . . ).
182
Recently, precision non-magnetic, calorimeter experiments in space (CALET [
27
],
ISS-183
CREAM [
28
], and DAMPE [
29
]) have also begun to provide important results.
184
There are excellent ongoing and proposed experiments that study cosmic rays at the
185
highest energies. This includes innovative experiments EUSO-SPB [
30
] and POEMMA [
31
],
186
as well as ground-based experiments – the Pierre Auger Observatory [
32
], H.E.S.S. [
15
],
187
LHAASO [
19
], KASCADE-Grande [
33
], TA [
34
], CTA [
35
], and others.
188
AMS is the first long duration (about two 11-year solar cycles), large acceptance precision
189
magnetic spectrometer to measure the sign and value of the charge, the momentum, the
190
rigidity, and the flux of elementary particles (positrons, electrons, antiprotons, protons),
191
nuclei, and anti-nuclei directly in space.
192 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57
Jour
nal
Pr
e-pr
oof
I. AMS ON THE ISS
193
As seen in Figure
2
, the AMS detector [
36
] consists of a permanent magnet and an
194
array of particle detectors to measure the velocity β = v/c, momentum P , charge Z, and
195
rigidity R = P/Z of traversing particles and nuclei. Within the magnet bore and above and
196
below the magnet are a total of 9 precision silicon tracker layers, L1 to L9. The Transition
197
Radiation Detector (TRD) is located at the top of the AMS. Above the magnet bore are
198
two orthogonal planes of Time of Flight counters (TOF), the Upper TOF, and below the
199
bore are another two orthogonal planes, the Lower TOF. The Anti-Coincidence Counters
200
(ACC or Veto), surround the tracker within the magnet bore. Below the Lower TOF is the
201
Ring Imaging Cherenkov counter (RICH) and below that the Electromagnetic Calorimeter
202
(ECAL). Each of these detector elements is described below.
203 TRD Upper TOF Tr a c k e r Lower TOF RICH ECAL 1 2 7-8 3-4 9 5-6 TRD identify e+, e -Silicon Tracker measure Z, P ECAL measure E of e+, e
-Upper TOF measure Z, E
Magnet identify ±Z, P
RICH measure Z, E
Lower TOF measure Z, E x
y z
B
ACC reject side particles
FIG. 2. The AMS detector showing the main elements and their functions. AMS is a TeV pre-cision, multipurpose particle physics magnetic spectrometer. It identifies particles and nuclei by their charge (Z), energy (E) and momentum (P ) or rigidity (R = P/Z), which are measured inde-pendently by the Tracker, TOF, RICH and ECAL. The ACC counters, located in the magnet bore, are used to reject particles entering AMS from the side. The AMS coordinate system, concentric with the magnet, is also shown. The x axis is parallel to the main component of the magnetic field and the z axis is pointing vertically.
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
Jour
nal
Pr
e-pr
oof
The separation of the TRD and the ECAL by a magnetic field ensures that most of the
204
secondary particles generated in the TRD are swept away by the magnet and do not enter
205
into the ECAL. In this way, the rejection powers of the TRD and the ECAL are independent.
206
Before launch to the ISS, AMS was tested extensively at the CERN test beam with
207
electrons, positrons, protons, and pions. Pion beams were used to simulate transition
radi-208
ation effects in the TRD for high energy protons. In total, more than 2000 combinations of
209
particles, energies, incident angles, and locations were tested.
210
AMS on the ISS has functioned reliably and the properties of the detector are continuously
211
monitored. Minute changes compared to the original calibration at the CERN accelerator
212
before launch are corrected in the data analysis. This ensures the quality and accuracy of
213
the data.
214
During more than 10 years of AMS construction, a large international group of physicists
215
have developed a comprehensive Monte Carlo simulation program for AMS. This program is
216
based on the geant4 package [
37
] and it is constantly being improved with the AMS data
217
obtained on the ISS. This Monte Carlo program simulates electromagnetic, hadronic, and
218
nuclear interactions of particles and nuclei in the material of AMS, namely:
219
1. Electromagnetic interactions including
220
• Ionization losses
221• Bremsstrahlung
222• Pair Production
223• Multiple and single Coulomb scattering
224
2. X-ray generation for the TRD detector
225
3. Cherenkov photon generation and propagation for the RICH detector
226
4. Elastic hadronic and nuclear scattering including
227
• Hadron elastic scattering
228
• Elastic nuclear scattering for ions
229
5. Inelastic hadronic and nuclear interactions from theoretical models tuned according
230
to our nuclear cross section measurements (see Section
VII
).
231
The trigger conditions and digitization of the signals are simulated according to the
232
measured characteristics of the electronics. The simulated events then undergo the same
233
reconstruction as used for the data.
234
The AMS coordinate system is concentric with the center of the magnet. The x-axis is
235
parallel to the main component of the magnetic field and the z-axis is parallel to the magnet
236 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57
Jour
nal
Pr
e-pr
oof
A. Permanent Magnet
241
The permanent magnet [
38
], shown in Figure
3
a, was flown on the AMS engineering
242
flight in 1998 and was described in detail in the AMS-01 Physics Report [
1
]. It is made of
243 244
64 high-grade Nd-Fe-B sectors assembled in a cylindrical shell 800 mm long with an inner
245
diameter of 1115 mm. This configuration produces a field of 1.4 kG at the center of the
246
magnet and negligible dipole moment and field leakage outside the magnet (see Figure
3
b).
247
This is important in order to eliminate the effect of torque on the ISS and to ensure safety
248
of astronauts. The detailed 3-dimensional field of the magnet was mapped in 2010 and is
249
shown in Figure
3
c. The field was measured in 120,000 locations to an accuracy of better
250
than 1%. Comparison with the measurements performed with the same magnet in 1997,
251
before the engineering flight AMS-01, shows that the field did not change within 1%, limited
252
by the accuracy of the 1997 measurement, as shown in Figure
3
d.
253
On the ISS, slight temperature induced changes in the field are constantly monitored and
254
corrected in the analysis.
255 1 2 3 4 5 6 7 89 10 1112 1314 15 Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α Α
b)
a)
Phi Angle (deg) 0 100 200 300 Radius (c m) 0 10 20 30 40 F ie ld (k G a u s s ) 0 0.5 1 1.5 2 0 0.5 1 1.5 2c)
Phi Angle (deg)
0 100 200 300 Rad ius (c m) 0 10 20 30 D e v ia ti o n (% ) 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
d)
FIG. 3. (a) The magnet being prepared for the AMS mission on the ISS. The magnet is 800 mm long with an inner diameter of 1115 mm. (b) The arrangement of the AMS magnet showing the field directions of the 64 permanent magnet sectors resulting in negligible dipole moment and field leakage outside the magnet. (c) The magnet field measured in 2010 at z = 0. (d) The deviation between the field measurements in 1997 and 2010. The Radius and Phi Angle are cylindrical coordinates in the x− y plane, such that Phi=0 corresponds to the x-axis.
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
Jour
nal
Pr
e-pr
oof
B. Tracker
256
The tracker, together with the magnet, accurately determines the trajectory and charge
257
Z of cosmic rays by multiple measurements of the coordinates (x and y) and energy loss. It is
258
composed of 192 ladders, each containing double-sided silicon sensors, readout electronics,
259
and mechanical support [
38
,
39
]. The AMS tracker has nine layers, L1–L9, and 196,608
260
readout channels. As shown in Figure
4
a, three planes of aluminum honeycomb with carbon
261
fiber skins are equipped with ladders on both sides. The layers on these planes are numbered
262
L3–L8. Another three planes are equipped with one layer of ladders each, numbered L1,
263
L2, and L9. As also indicated in Figure
4
a, L1 is located on top of the TRD, L2 is above
264
the magnet, and L9 is between the RICH and the ECAL. L9 covers the ECAL acceptance.
265
L2–L8 constitute the inner tracker. The total lever arm of the tracker from L1 to L9 is 3.0
266
m. From the trajectory and the magnetic field map, the tracker measurement directly yields
267
the rigidity or momentum per unit charge, R = P/Z.
268
Positions of the planes of the inner tracker are held stable by a special carbon fiber
269
structure [
1
]. It is monitored using 20 IR laser beams, which penetrate through all planes of
270
the inner Tracker and provide micron level accuracy position measurements. The positions
271
of L1 and L9 are aligned using cosmic ray protons every 2 minutes. As seen in Figures
4
b
272
and
4
c, they are stable to 2–3 µm.
273 L1 L5 L6 L3 L4 L7 L8 L9 L2 L1 L2 L9 L3−L4 L5−L6 L7−L8 -50 0 50 100 m) μ Residual in Y ( -20 -10 0 10 20 -20 -10 0 10 20 En tr ie s En tr ie s 0 1000 2000 3000 4000 m μ = 2.2 σ Ma y J u l Se p N o v J a n Ma r Ma y J u l Se p N o v J a n Ma r Ma y J u l Se p N o v J a n Ma r Ma y J u l Se p N o v J a n Ma y Ma r J u l Se p N o v J a n Ma r Ma y J u l Se p N o v J a n Ma r Ma y J u l Se p N o v J a n Ma r Ma y m ) μ R e s id u a l in Y ( m ) μ R e s id u a l in Y ( -50 0 50 100 0 1000 2000 3000 σ= 2.3 μm 2011 2012 2013 2014 2015 2016 2017 2018 b) Layer 1 c) Layer 9 a)
FIG. 4. (a) The 9 layers of the AMS silicon tracker and their locations within the detector. The alignment stability of (b) Layer 1 and (c) Layer 9 over seven years.
1. Tracker coordinate measurements
274 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57
Jour
nal
Pr
e-pr
oof
of the rigidity (or momentum) [
40
]. This provides spatial resolution of 5
− 10µm in the
279
bending plane and 13
− 20µm in the non-bending plane.
280
We present the method of determination of the y coordinate. The method of
determina-281
tion of the x coordinate is similar.
282
When a charged particle crosses a layer of the silicon tracker, its coordinate d (either
283
in x or y) is determined by taking the ratio between the signals induced on the two strips
284
between which the particle passed (see Figure
5
).
285
The amplitude of the induced signals should be proportional to Z
2. For high Z, the
286
amplitudes gradually saturate and become non-linear. This causes coordinate resolution
287
degradation.
288
We have developed an optimal technique to correct for the non-linear effect. As illustrated
289
in Figure
5
, we identify a function K(d) which will restore the linearity of the amplitudes
290
by taking into account measured amplitudes A
0, A
1and the corrected position d:
291
d = 1/(1 + (A
0+ K(d))/A
1).
(1)
{
{
A
1
A
0
K(d)
Corrected
Uncorrected
0
1
d
FIG. 5. Schematic of particle coordinate measurement in the silicon tracker. The solid red arrow indicates the point where the particle intersected the silicon sensor, the two adjacent strips are positioned at 0 and 1 in “strip” units. The maximum amplitude is defined as A0 (at 0) and the
next largest adjacent strip as A1 (at 1). For an ideal tracker, the ratio between amplitude A0
and A1 gives the particle coordinate, d = 1/(1 + A0/A1). To correct for the non-linear effect, the
correction function K(d) is added to the measured non-linear amplitude A0. The dashed arrow
shows the uncorrected coordinate measurement. The corrected coordinate is at its true position within the measured accuracy of few microns.
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
Jour
nal
Pr
e-pr
oof
To determine K(d), we used the fact that cosmic rays are uniform and isotropic, so for
292
an ideal tracker, without nonlinearity, we should see a uniform event position density (see
293
Figure
6
a), while in the non-linear case the event density is distorted (see Figure
6
b).
294
The function K(d) was found to be different for different Z. Figure
6
c shows the event
295
position density distribution for carbon nuclei before and after correction. With this
correc-296
tion, the observed coordinate accuracy in the bending plane is 6.5 µm for helium, 5.1 µm
297
for carbon, and 6.3 µm for oxygen.
298
0.5
-0.5
0
d
Tracker linear response
Tracker non-linear response
Uniform Event Position Density
Distorted Event Position Density
0.5
-0.5
0
d
a)
b)
Position -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 Ev e n ts D e n s ity 0 0.01 0.02 0.03 0.04A1
A0
A1
c)
FIG. 6. (a) The uniform event position density for an ideal tracker. (b) The distorted event position density for the tracker with non-linear amplitude response. (c) The event position density before correction (black curve) and after correction (red curve) for carbon. The strips are positioned at 0 for the strip with the amplitude A0, and at 1 or -1 for A1.
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57
Jour
nal
Pr
e-pr
oof
Figure
7
shows the coordinate measurement accuracy for nuclei from Z = 2 to Z = 26
299
for two different track geometries: L1–L9 and L1–L8. Note that, the L1–L9 geometry has
300
tracks with almost normal incidence and therefore has the best coordinate resolution, while
301
the L1–L8 geometry has tracks with larger inclination angles. Due to the design of the
302
tracker readout amplifier, the maximum non-linearity occurs for Z
∼ 9.
303 2 4 6 8 10 12 14 16 m] μ [ σ R e si d u a ls 3 4 5 6 7 8 9 10 11 L1-L9 Geometry Z 5 10 15 20 25 m] μ [ σ R e si d u a ls 3 4 5 6 7 8 9 10 11 L1-L8 Geometry a) b) Z
FIG. 7. AMS tracker residuals σ as functions of nuclei charge Z for (a) the L1–L9 geometry and (b) the L1–L8 geometry. The residuals were obtained by comparison of the differences of the coordinates measured in layers L3 or L5 to those obtained from the track fit using the measurements from L1, L2, L4, L6, L7, and L8 in the rigidity range R > 50 GV. Due to the design of the tracker readout amplifier, the maximum non-linearity occurs for Z ∼ 9.
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
Jour
nal
Pr
e-pr
oof
2. Tracker rigidity measurement
304
Together, the tracker and the magnet determine the rigidity R of charged cosmic rays
305
by measuring the particle trajectory in the magnetic field. To find the particle trajectory
306
we use a track-finding algorithm based on cellular automaton for finding the track segments
307
and then constructing the track, as illustrated in Figure
8
. This improves the track finding
308
efficiency and rejection of spurious hits in the detector [
41
].
309
FIG. 8. Schematics of the track-finding algorithm. (a) Event with original hits. (b) Event with all track segments. (c) Correct segments chosen based on minimization of angles between segments. (d) Chose one track as the result of using a χ2-like track quality estimator.
This algorithm is particularly important for heavy nuclei events, in which additional hits
310
or track segments are often present due to delta-ray generation and nuclei interactions with
311
the tracker materials.
312
Once a track has been found, its rigidity is determined using a track-fitting algorithm
313
based on the Kalman filtering technique. It accurately accounts for energy losses and
mul-314
tiple scattering by charged particles [
42
]. For low rigidities (< 20 GV) the ∆R/R is 0.1.
315
This is particularly important for the measurements of nuclear isotopes. The maximum
de-316
tectable rigidity (MDR) with this algorithm is 2.0 TV for protons, 3.2 TV for helium, 3.7 TV
317
for carbon, 3.4 TV for oxygen, and 3.7 TV for iron.
318
Test beam data are important in understanding the tracker performance. Figure
9
a
319
shows the comparison between data and the Monte Carlo simulation of the inverse rigidity
320
measured by the tracker for 400 GeV/c protons from the CERN test beam. As seen, the
321
Monte Carlo simulation describes not only detector resolution effects (the central part of
322
the distribution) but also the effects of interactions with the detector materials including
323
multiple, large angle, elastic, and quasi-elastic scattering (the tails of the distribution).
324
Figure
9
shows that the agreement between the data and the Monte Carlo simulation extends
325
over five orders of magnitude. To study the tracker performance beyond the test beam
326
momentum (400 GeV/c proton beam), we used the data from the ISS to compare the rigidity
327 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57
Jour
nal
Pr
e-pr
oof
-0.02
-0.01
0
0.01
0.02
Ev
e
n
ts
10
210
310
410
510
610
400 GeV/c Test Beam
400 GeV/c Simulation
1
Rigidity
400
[GV ]
1
-1]
-1[TV
L2−9R
1
-
L1−8R
1
-20
-15
-10
-5
0
5
10
15
20
E
v
e
n
ts
1
10
210
310
1130 < R < 1800 GV
p data
p MC
a)
b)
FIG. 9. (a) Comparison between data and the Monte Carlo simulation of the inverse rigidity measured by the tracker for 400 GV test beam protons. As seen, the agreement between the data and the Monte Carlo simulation extends over five orders of magnitude. (b) The difference of the inverse rigidities measured with the upper (Layers L1–L8) and the lower (Layers L2–L9) parts of the tracker for the cosmic ray proton data collected on the ISS and for the Monte Carlo simulation in the rigidity range [1130-1800] GV.
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
Jour
nal
Pr
e-pr
oof
3. Determination of tracker absolute rigidity scale and alignment
332
In AMS, for all Z, the largest systematic error in the determination of the fluxes at the
333
highest energies is due to the uncertainty of the absolute rigidity scale. The AMS tracker
334
alignment and the absolute rigidity scale determination were performed before launch using
335
the CERN test beam data, as shown in Figure
9
a. Vibrations and accelerations on the ISS
336
due to movement of solar arrays, the attitude change, docking and undocking of the visiting
337
vehicles, astronaut activities, and during the AMS launch into space as well as outgassing
338
of the carbon fiber supporting structure in vacuum, may change the ladder positions of
339
the inner tracker at the sub-micron level, and therefore cause shifts in the absolute rigidity
340
scale. Note, a coherent shift in the inner tracker layers of less than 0.5 microns is sufficient
341
to create an absolute rigidity scale shift of 10% at 1 TV.
342
The in-flight rigidity scale shift s and its uncertainty were obtained by the comparison of
343
the inverse absolute rigidity 1/
|R|, measured by the tracker, with the inverse energy 1/E,
344
measured by the electromagnetic calorimeter, for positron events and electron events [
43
].
345
• First, the electron events and positron events were split into 72 energy bins from 2 to
346
300 GeV, with bin widths chosen according to the calorimeter energy resolution.
347
• Next, probability density function (PDF) for each bin were calculated from the 1/|R|−
348
1/E electron distributions. Figure
10
shows the 1/
|R| − 1/E distribution for electron
349
events and its parametrization together with the positron events in the 11 to 13 GeV
350
bin.
351
• Each PDF is then parametrized by the sum of a Gaussian function and an
exponen-352
tially modified Gaussian function.
353
• The bin-by-bin PDF distributions are parametrized as functions of energy.
354
• The resulting energy dependent PDF, f(1/|R| − 1/E, E), is then used in an unbinned
355
likelihood fit of the rigidity scale shift parameters, namely s
+for positrons and s
−for
356
electrons.
357
• The likelihood is defined as
P
+logf (1/
|R| + s
+− 1/E, E) +
P
−logf (1/
|R| + s
−−
358
1/E, E), where the summations include all positron events or electron events,
respec-359
tively.
360
• The absolute rigidity scale shift is then evaluated as s = (s
+− s
−)/2.
361 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57
Jour
nal
Pr
e-pr
oof
]
-1[TV
E
1
-
|R|
1
-20
-10
0
10
20
30
40
Eve
n
ts
100
11<E<13 GeV
+e
11<E<13 GeV
-e
parameterization
-e
11<E<13 GeV
+e
11<E<13 GeV
-e
parameterization
-e
20
200
FIG. 10. The 1/|R| − 1/E distribution for the energy bin 11 to 13 GeV for electrons (red data points) and positrons (blue data points). The distribution for electrons is parametrized by the sum (red curve) of a Gaussian and an exponentially modified Gaussian with a χ2/d.o.f. = 65/68.
As mentioned above, various effects at launch and in space may modify rigidity scale
362
and bias flux measurements. To determine the time-dependent rigidity shift correction, the
363
electron-positron data from the first 7 years of operations is divided into four time intervals.
364
Figure
11
shows the time dependence of the rigidity scale correction for the period from May
365
2011 to May 2018. With this time dependent correction, the accuracy of the rigidity scale
366
shift is found to be within 0.033 TV
−1or 3% at 1 TV, limited mostly by available positron
367
statistics.
368 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63Jour
nal
Pr
e-pr
oof
]
-1-0.3
-0.2
-0.1
0
0.1
0.2
0.3
Year
2013
2015
2017
R
ig
id
ity
s
c
a
le
c
o
rr
e
c
ti
o
n
[
T
V
FIG. 11. The time dependence of the rigidity scale correction for the period from May 2011 to May 2018.
The positions of L1 and L9 are dynamically aligned every 2 minutes according to
extrap-369
olations from the inner tracker (layers L2 to L8). Therefore, the rigidity scale of the full
370
span tracker with layers from L1 to L9 follows the rigidity scale of the inner tracker. To
371
verify this, the difference in the rigidity scale shifts between the full span tracker and the
372
inner tracker, s(1/R
19)
− s(1/R
28), has been estimated using high energy cosmic ray proton
373
and helium events. Both proton and helium samples yield similar results. The difference
374
is found to be s(1/R
19)
− s(1/R
28) =
−0.019 ± 0.011 TV
−1independent of rigidity [
43
].
375
The corresponding small shifts in the L1 and L9 positions were corrected and the error of
376
0.011 TV
−1is added in quadrature to the total error of the rigidity scale.
377
Overall, the tracker rigidity scale was measured with an accuracy of
±0.034 TV
−1(i.e
378
quadratic sum of 0.033 TV
−1and 0.011 TV
−1). This corresponds to the determination of a
379
coherent displacement of layers L2–L8 by less than 0.2 microns.
380
To verify that after these corrections rigidity scale is time independent, we have studied
381
the time dependence of the measured fluxes. Figure
12
shows the ratio of the proton and
382 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57
Jour
nal
Pr
e-pr
oof
11
Rigidity[GV]
1 10 102 103
Ratios of Proton Fluxes
0.85 0.9 0.95 1 1.05 1.1 1.15 May 2011-Jan 2013 Feb 2013-Oct 2014 Nov 2014-Jul 2016 Aug 2016-May 2018 Rigidity[GV] 10 102 103
Ratios of Helium Fluxes
0.85 0.9 0.95 1 1.05 1.1 1.15 May 2011-Jan 2013 Feb 2013-Oct 2014 Nov 2014-Jul 2016 Aug 2016-May 2018 a) b)
FIG. 12. The ratios of the (a) proton and (b) helium fluxes measured for each of the four 21-month periods to the flux measured over 7 years. At low rigidities, the fluxes are time dependent due to solar modulation. At high rigidities, above 100 GV, the fluxes are consistent within measurement errors.
4. Charge-sign identification
387
Charge–sign identification is a critical property of the spectrometer. It allows AMS
388
to distinguish negatively charged cosmic rays from positively charged cosmic rays. It is
389
quantified by the amount of charge confusion. For example, in the analysis of positrons
390
it is important to distinguish positrons from charge-confusion electrons, i.e. those electrons
391
reconstructed with the positive charge sign due to the finite tracker resolution or interactions
392
in the detector materials. To this end, a charge confusion estimator Λ
eCC
is defined using the
393
Boosted Decision Trees technique [
44
,
45
]. The estimator combines several measurements:
394
the ratio of the energy from the calorimeter to the momentum from the tracker, E/p, the
395
track χ
2/d.o.f., momenta reconstructed with different combinations of tracker layers, the
396
number of hits in the vicinity of the track, and the charge measurements in the TOF and in
397
the tracker. With this method, positrons, which have Λ
eCC
∼ +1, are efficiently separated
398
from charge confusion electrons, which have Λ
eCC
∼ −1.
399
After selection of a sample of positron and electron events with the TRD and ECAL,
400
the charge confusion in data is obtained using the template fitting technique with the Λ
e CC401
distribution [
45
]. From the fit to the positive rigidity sample, we obtain the number of
402
positron N
e+events, charge confusion electron background N
ec.c.−events, and proton
back-403ground events. From the fit to the negative rigidity sample, we obtain the number of
elec-404
tron N
e−events, charge confusion positron background N
ec.c.+events, and proton background
405events. The charge confusion fraction in the electron data is calculated as N
c.c.e−
/(N
ec.c.−+N
e−)
406and similarly for positrons.
407
The charge confusion in the Monte Carlo simulation is directly calculated as the fraction
408
of electrons being reconstructed with positive rigidity after the same selection cuts as used
409
in data. As an illustration, the comparison between the electron charge confusion fraction
410
in the data and in the Monte Carlo simulation is presented in Figure
13
. As seen, the charge
411
confusion is well reproduced by the Monte Carlo simulation. The charge confusion fraction
412
is <8% in the energy range up to 1 TeV.
413 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
Jour
nal
Pr
e-pr
oof
Energy [GeV]
1
10
10
2
10
3
C
h
a
rg
e
c
o
n
fu
s
io
n
fr
a
c
ti
o
n
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
MC
Data
FIG. 13. The charge confusion fraction as a function of energy measured with electrons. The charge confusion fraction is <8% in the energy range up to 1 TeV. As seen, the charge confusion is well reproduced by the Monte Carlo simulation.
Similar methods are used to differentiate antiprotons and anti-deuterons from proton and
414
deuteron backgrounds.
415 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57Jour
nal
Pr
e-pr
oof
5. Tracker charge measurement
416
The nine tracker layers independently measure the charge
|Z| of cosmic rays. The
ion-417
ization energy losses deposited in the silicon are proportional to Z
2and they are measured
418
with strips on both the x and y sides of a silicon sensor.
419
The energy loss deposition is collected by several strips on both x and y sides. The
420
relative signal on either side is related to the transverse distance d between the nearest strip
421
and the location where the particle crosses the silicon, as shown in Figure
14
. The strip
422
amplitudes are first corrected for electronics gain factors and then for the dependence on
423
the particle position d and inclination angle θ. Details of this technique as well as the effects
424
of non-linear electronics response for high Z nuclei are described in Ref. [
46
].
425
COG
-2
-1
0
1
2
ADC
0
100
200
300
400
1Signal Response of strips in Tracker
!
Angle of inclination
Position d
A
4
A
2
A
0
A
1
A
3
A
#
($=1-5) are the strips with five largest signal responses.
A
1
is called “
seed strip
”
FIG. 14. Schematic of ionization energy deposition signals per strip, in ADC counts, at the incident particle position d (defined in Eq.1, see also Figure 5) and the inclination angle θ of the particle (red arrow). A0 is the strip with the highest amplitude (that is, the strip nearest to the crossing
point), the next highest adjacent strip A1, the other adjacent strip A2, etc...
426 427 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
Jour
nal
Pr
e-pr
oof
The charge measurement is obtained by combining the charge determined by both x and
428
y strips. Figure
15
shows the charge resolution of tracker layers L2–L8 for nuclei from Z = 1
429
to Z = 28. This resolution enables us to perform the precision measurement of all nuclei
430
fluxes up to and beyond nickel (Z = 28).
431
Z
Charge
0
5
10
15
20
25
30
Charge Resolution
∆
Z
/Z
0
0.01
0.02
0.03
0.04
0.05
He
C
Ne
Si
Ca
Fe Ni
FIG. 15. The inner tracker (layers L2–L8 combined) charge resolution ∆Z/Z. The solid line is to guide the eye.
432 433 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57
Jour
nal
Pr
e-pr
oof
Figure
16
shows the charge measured simultaneously by the TOF (see section
I D
) and
434the tracker.
435ToF
Ch
arge
0
5
10
15
20
25
30
Tracke
r Charg
e
0
5
10
15
20
25
30
E
ve
n
ts
2
10
3
10
4
10
5
10
6
10
7
10
8
10
9
10
H Li He Be B C O Ne Na Mg Al Si P S Cl Ar K Ca Sc Ti V Cr Mn Fe Co Ni F NFIG. 16. The charge measurement by the TOF (see Figure22) and the tracker.
436 437 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
Jour
nal
Pr
e-pr
oof
C. Transition Radiation Detector (TRD)
438
The Transition Radiation Detector (TRD) [
47
] is located at the top of the AMS as shown
439
in Figure
2
. Its main purpose is to identify electrons and positrons by transition radiation
440
while rejecting protons at a level of 10
3. The TRD also provides an independent tracking
441
capability and determination of the charge value of the nuclei by measuring their rate of
442
energy loss (dE/dx).
443
The TRD consists of 5248 proportional tubes of 6 mm diameter with a maximum length
444
of 2 m arranged side-by-side in 16–tube modules. The 328 modules are mounted in 20
445
layers. As shown in Figure
17
a, the assembly of the TRD layers is supported in a tapered
446
octagonal carbon fiber structure with a very low coefficient of thermal expansion. Such a
447
structure ensures the minimum relative movement of the TRD elements with the variation
448
of the ambient temperature.
449
As shown in Figure
17
b, each layer is interleaved with a 20 mm thick fiber fleece
ra-450
diator, LRP375BK, with a density of 0.06 g/cm
3. There are twelve layers of proportional
451
tubes along the y axis located in the middle of the TRD and, along the x axis, four layers
452
located on top and four on the bottom. The tubes are filled with a 90:10 Xe:CO
2gas
mix-453
ture. Xenon efficiently captures the transition radiation X-rays generated in the radiator
454
layers. CO
2ensures stable operation of the proportional tubes. An anode wire in each straw
455
tube measures the signal from the resulting ionization due to the captures of the transition
456
radiation photons as well as the ionization signal of the traversing charged particle (see
457