Etude et impact du bruit de fond corrÃ©lÃ© pour la mesure de l'angle ...

More documents

Recommendations

Info

110 5. Correlated background can be estimated comparing tagged and un-tagged sample in the high energy tail of the prompt spectrum, assuming they would remain unchanged at low energy < 12 MeV. The estimation of the tagging e ciencies represent the main limitation of these techniques. However, IV and OV are independent detectors based on di↵erent technologies. They would provide independent estimation of correlated background which, if in agreement, would validate the assumption made on the tagging e ciency. IV tagging tel-00821629, version 1 - 11 May 2013 Since FN are generated with high multiplicity, some of them are expected to deposit energy in both the ID and IV. The DAQ system is designed in such a way that both ID and IV detectors are read-out whatever detector causes the trigger. In case of ID trigger, the IV detection threshold is then lowered to ⇠ 1 MeV, making the IV sensitive to FNs via proton recoil or captures on H. Defining a tagging condition in the IV would then allow to select a sample of FN which would provide the spectral shape in [0.7, 12.2] MeV energy region. Given the small angular acceptance by which the SM enter the ID through the chimney, IV is expected to tag SM with smaller e ciency. A di↵erent technique to infer the SM spectral shape is to take advantage of the Michel electron spectra, which extend up to ⇠ 60 MeV, and it is uncorrelated from the energy deposited by its parent muon. A pure sample of SM could then be obtained selecting the delayed Michel electron at high energy, > 20 MeV. In this energy region, ¯⌫ e and FN contaminations are expected to be negligible, since their delayed events deposit energy by Gd capture between 6 MeV and 12 MeV. OV tagging The OV is a suitable detector to identify correlated background by tagging the parent muon. The tagging is performed by time coincidences between an OV hit and the prompt-like energy deposition in the ID. The SM component is expected to be tagged with high e ciency since the muon enter in the detector through the chimney which is surrounded by OV scintillator strips. A smaller tagging e ciency is expected for FN since the parent muon could interact far from the detector and miss the OV. Pulse shape discrimination A pulse shape discrimination (PSD) technique is also been investigated to perform event tagging. PSD is expected to tag FNs via the di↵erent scintillation shape produced by the recoil protons, with respect to electrons and positrons. The distribution of the time of each pulse in each event is compared to a reference distribution, obtained with 137 Cs calibration source,
5.2 High energy analysis 111 in a likelihood approach. Despite the DC liquid scintillator has not been optimised for PSD, preliminary results are promising [73]. More studies are expected in the future. Analysis based on the extrapolation of high energy tail and both IV and OV tagging have been performed for the second DC publication. Those di↵erent approaches are found to provide compatible results and will be described in the following sections. 5.2 High energy analysis tel-00821629, version 1 - 11 May 2013 Correlated background sample is performed using the ¯⌫ e selection cut summarised in Sec. 4.2.1, extending prompt energy cut from 12 MeV to 30 MeV. Since reactor ¯⌫ e spectrum ends at about 12 MeV, above such energy a pure sample of correlated background is obtained. The prompt and delayed vertex distributions are shown in Fig. 5.1. The black points represent events in [0.7, 30] MeV, including ¯⌫ e candidates, while red points shown the vertex distribution of the correlated background sample, with prompt energy in [12, 30] MeV. The events uniformly distributed among the detector are expected to be dominated by FN while the excess of events below the detector chimney are expected to be more likely due to SM. Since the reconstruction algorithm has been designed to reconstruct events within the bu↵er volume, interaction vertices in the chimney are artificially pushed down toward the detector center. The true SM vertex distribution is then expected to be slightly above the observed one. The definition of a fiducial volume to separate FN from SM would then introduce systematic uncertainty related to the vertex reconstruction algorithm. In order to avoid such systematic uncertainty no fiducial volume is defined. The prompt spectrum in [0.7, 30] MeV is shown in Fig. 5.2. The correlated background spectrum shows a rather flat distribution for energies > 12 MeV, while its spectral shape in the low energy region could not be inferred since the ¯⌫ e are dominant. A linear fit in the high energy region, [12, 30] MeV, provides a rather flat spectrum with small negative slope of ( 0.08 ± 0.04)/(0.25 MeV) 2 . The extrapolation of the linear fit to the low energy region provide a first estimation of the correlated background rate of (0.75 ± 0.12) cpd. The same method has been adopted by Daya Bay [33] and RENO [31], but as already mentioned, the accuracy of this estimation depends on the validity of the extrapolation of the spectral shape. In order to evaluate FN and SM background independently with dedicated analysis, the two component have to be separated. Since FN and SM present di↵erent time correlation between prompt and delayed events, a pure and unbiased sample of each component is obtained by a cut on the t between
Page 1 and 2:
UNIVERSITÉ DENANTES FACULTÉ DES S
Page 3 and 4:
tel-00821629, version 1 - 11 May 20
Page 5 and 6:
Abstract tel-00821629, version 1 -
Page 7 and 8:
Acknowledgments Since so many peopl
Page 9 and 10:
Contents 1 Introduction 13 tel-0082
Page 11 and 12:
CONTENTS 11 tel-00821629, version 1
Page 13 and 14:
Chapter 1 Introduction tel-00821629
Page 15 and 16:
15 tel-00821629, version 1 - 11 May
Page 17 and 18:
Chapter 2 Neutrino physics tel-0082
Page 19 and 20:
2.2 Neutrino history in brief 19 ma
Page 21 and 22:
2.3 Neutrino oscillation, the theor
Page 23 and 24:
2.4 Measuring neutrino oscillation
Page 25 and 26:
3-flavour oscillations The dominant
Page 27 and 28:
Page 29 and 30:
After the Sun, the other main natur
Page 31 and 32:
Page 33 and 34:
Page 35 and 36:
2.5 The problem with the neutrino m
Page 37 and 38:
2.6 Summary and open questions 37 s
Page 39 and 40:
In the next future, reactor experim
Page 41 and 42:
Chapter 3 The Double Chooz experime
Page 43 and 44:
3.1 The Double Chooz detector 43 to
Page 45 and 46:
3.1 The Double Chooz detector 45 te
Page 47 and 48:
3.1 The Double Chooz detector 47 it
Page 49 and 50:
3.1 The Double Chooz detector 49 te
Page 51 and 52:
3.2 The detector read-out 51 40 m o
Page 53 and 54:
3.3 The online system 53 IV trigger
Page 55 and 56:
3.3 The online system 55 tions as s
Page 57 and 58:
3.3 The online system 57 tel-008216
Page 59 and 60: 3.4 The calibration system 59 tel-0
Page 61 and 62: loosing their kinetic energy 3 and
Page 63 and 64: 3.5 Reactor neutrino flux simulatio
Page 69 and 70: 3.8 Data reconstruction 69 A dedica
Page 71 and 72: Regarding the pulse information, th
Page 73 and 74: 3.8 Data reconstruction 73 tel-0082
Page 75 and 76: 3.8 Data reconstruction 75 Gain (ch
Page 77 and 78: 3.9 Conclusions 77 tel-00821629, ve
Page 79 and 80: Chapter 4 Measurement of ✓ 13 wit
Page 81 and 82: 4.1 Data sample 81 create a signal
Page 83 and 84: 4.2 ¯⌫ e signal selection 83 Liv
Page 85 and 86: 4.2 ¯⌫ e signal selection 85 Pro
Page 87 and 88: 4.2 ¯⌫ e signal selection 87 Neu
Page 89 and 90: 4.3 Backgrounds 89 energy spectrum
Page 91 and 92: 4.3 Backgrounds 91 tel-00821629, ve
Page 93 and 94: 4.4 Selection e ciencies and system
Page 95 and 96: 4$&14+#':"'$"$468'#21%$;' ! $"$468'
Page 97 and 98: 4.4 Selection ! e ciencies and syst
Page 99 and 100: 4.5 Data analysis summary 99 4.4.3
Page 101 and 102: 4.6 Final fit and measurement of
Page 103 and 104: 4.6 Final fit and measurement of
Page 105 and 106: 4.7 Summary and Conclusions 105 tel
Page 107 and 108: Chapter 5 Correlated background tel
Page 109: 5.1 Measuring correlated background
Page 113 and 114: 5.2 High energy analysis 113 tel-00
Page 115 and 116: 5.3 OV Tag analysis 115 The t cut v
Page 117 and 118: 5.4 Fast Neutrons analysis 117 bigg
Page 119 and 120: 5.4 Fast Neutrons analysis 119 5.4.
Page 121 and 122: 5.4 Fast Neutrons analysis 121 tel-
Page 123 and 124: 5.4 Fast Neutrons analysis 123 esti
Page 125 and 126: 5.4 Fast Neutrons analysis 125 Corr
Page 129 and 130: 5.4 Fast Neutrons analysis 129 Prom
Page 131 and 132: 5.4 Fast Neutrons analysis 131 lect
Page 133 and 134: 5.4 Fast Neutrons analysis 133 mari
Page 139 and 140: 5.5 Stopping Muon analysis 139 5.5.
Page 141 and 142: 5.5 Stopping Muon analysis 141 tel-
Page 143 and 144: 5.5 Stopping Muon analysis 143 tel-
Page 145 and 146: 5.5 Stopping Muon analysis 145 Sour
Page 147 and 148: 5.5 Stopping Muon analysis 147 belo
Page 149 and 150: 5.5 Stopping Muon analysis 149 The
Page 151 and 152: 5.6 Estimation of correlated backgr
Page 157 and 158: 5.7 Conclusion 157 5.7 Conclusion t
Page 159 and 160: Chapter 6 Conclusions tel-00821629,
Page 161 and 162:
161 • The SM analysis could be im
Page 163 and 164:
Chapter 7 Contributions to Double C
Page 165 and 166:
165 from this study and from [64] f
Page 167 and 168:
List of Figures tel-00821629, versi
Page 169 and 170:
LIST OF FIGURES 169 tel-00821629, v
Page 171 and 172:
Page 173 and 174:
Page 175 and 176:
Page 177 and 178:
Page 179 and 180:
List of Tables 3.1 Mean energy rele
Page 181 and 182:
Bibliography [1] [2] tel-00821629,
Page 183 and 184:
BIBLIOGRAPHY 183 [31] J.K. Ahn et a
Page 185 and 186:
BIBLIOGRAPHY 185 [61] W. Hampel et
Page 187:
BIBLIOGRAPHY 187 [92] A. Stueken et
show all

Etude et impact du bruit de fond corrÃ©lÃ© pour la mesure de l'angle ...

Create successful ePaper yourself

Delete template?

Save as template?