Search for new phenomena in top-antitop quarks final states with additional heavy-flavour jets ... with the atlas detector

This book reports on the search for a new heavy particle, the Vector-Like Top quark (VLT), in the Large Hadron Collider (LHC) at CERN. The signal process is the pair production of VLT decaying into a Higgs boson and top quark (TT→Ht+X, X=Ht, Wb, Zt). The signal events result in top–antitop quarks final states with additional heavy flavour jets. The book summarises the analysis of the data collected with the ATLAS detector in 2015 and 2016. In order to better differentiate between signals and backgrounds, exclusive taggers of top quark and Higgs boson were developed and optimised for VLT signals. These efforts improved the sensitivity by roughly 30%, compared to the previous analysis. The analysis outcomes yield the strongest constraints on parameter space in various BSM theoretical models. In addition, the book addresses detector operation and the evaluation of tracking performance. These efforts are essential to properly collecting dense events and improving the accuracy of the reconstructed objects that are used for particle identification. As such, they represent a valuable contribution to data analysis in extremely dense environments. Supervisor’s Foreword 6 Parts of this thesis have been published in the following journal articles 8 Acknowledgements 9 Contents 11 Acronyms 16 1 Introduction 18 1.1 Standard Model 18 1.1.1 Electroweak Theory 20 1.1.2 The Higgs-Englert-Brout Mechanism 21 1.1.3 Quantum Choromo-Dynamics 24 1.1.4 Experiment Results on the Standard Model 25 1.2 Beyond the Standard Model 30 1.2.1 Extra Dimension 30 1.2.2 Compositeness 32 1.3 Signatures of the Physics Beyond the Standard Model 33 1.3.1 Vector-Like Quark 34 1.3.2 Four-Top-Quark Production 38 References 39 2 The ATLAS Experiment at the Large Hadron Collider 42 2.1 Large Hadron Collider 42 2.1.1 Accelerators 43 2.1.2 The Design of the Large Hadron Collider 43 2.2 ATLAS Detector 46 2.2.1 Coordinate System 47 2.2.2 Magnet System 48 2.2.3 The Inner Detector 49 2.2.4 The Calorimeters 53 2.2.5 Muon Spectrometer 56 2.2.6 Trigger System 57 2.2.7 Luminosity Detector 59 2.2.8 Operation and Data Quality 60 2.3 Studies of Monitoring Pixel Detector 61 2.3.1 Hit Information 61 2.3.2 Pileup Dependency of Hit Occupancy 62 2.3.3 Synchronisation Error Rate 65 2.3.4 Summary 67 References 67 3 Reconstruction of Physics Objects 69 3.1 Tracks 69 3.1.1 Study of Impact Parameter Resolutions 72 3.2 Primary Vertices 77 3.3 Leptons 78 3.3.1 Electrons 78 3.3.2 Muons 80 3.4 Jets 82 3.5 Large Radius Jets 84 3.6 b-Tagging 85 3.7 Missing Transverse Energy 89 3.8 Overlap Removal 91 References 91 4 Data and Event Simulations 94 4.1 Data Sample 94 4.2 Simulation of pp Collisions 95 4.2.1 Event Generation 95 4.2.2 Detector Simulation 97 References 97 5 Signal and Background Modeling 99 5.1 Signal Modeling 99 5.1.1 Vector-Like Quark Pair Production 100 5.1.2 Four-Top-Quark Production 102 5.2 Background Modeling 102 5.2.1 tbart+jets Background 103 5.2.2 W/Z+jets Background 103 5.2.3 Single-Top Backgrounds 104 5.2.4 Diboson Background 104 5.2.5 tbart+V and tbart+H Backgrounds 104 5.2.6 Multijet Background 104 5.3 Common Modelling and Simulation to All the Signals and Backgrounds 107 5.4 Tag-Rate-Function Method 107 5.5 Appendix 109 5.5.1 Signal Chirality Study 109 5.5.2 Validation Plots for the Multijet Background Estimation 122 5.5.3 TRF Closure Tests 138 References 143 6 Optimisations of Heavy Object Tagging and Event Categorisation 146 6.1 Methodology of the Optimisation 146 6.2 Optimisation of Higgs Tagging 147 6.2.1 Determination of Higgs Tagging Requirements 148 6.2.2 Performance of the Higgs Boson Tagger 153 6.3 Optimisation of the Top Quark Tagger 161 6.3.1 Determination of Top Tagging Requirements 161 6.3.2 Performance of the Top Tagger 162 6.4 Summary for Determination of the Taggers 167 6.5 Optimisation of Event Categorisation 167 6.5.1 Search Regions 168 6.5.2 Validation Regions 172 6.5.3 Summary for Determination of the Event Categorisation 175 6.6 Appendix 178 6.6.1 Auxiliary Materials for the Optimisation 178 6.6.2 Kinematic Variables in Validation Regions 179 References 187 7 Analysis Strategy 188 7.1 Preselection 188 7.2 Event Classification 191 7.3 Signal-to-Background Discrimination 193 7.4 Blinding Strategy 196 7.5 Data-to-MC Comparison in Blinded Regions 199 7.6 Appendix 200 7.6.1 Blind Cuts on meff 200 7.6.2 Data/MC Comparison in Blinded Regions 200 Reference 210 8 Systematic Uncertainties 211 8.1 Luminosity 211 8.2 Pileup 211 8.3 Physics Objects 213 8.3.1 Lepton 213 8.3.2 Jets 213 8.3.3 Missing Transverse Energy 214 8.3.4 Large-R Jets 215 8.3.5 Flavor Tagging 215 8.4 tbart+jets Background 215 8.5 W/Z+jets Background 217 8.6 Single Top Backgrounds 217 8.7 Other Backgrounds 218 8.8 Multijet Background 218 References 219 9 Statistical Analysis 221 9.1 Likelihood Function and Profile Likelihood Ratio 222 9.2 Confidence Level 225 References 227 10 Results 228 10.1 Fit Procedure 228 10.2 Expected Fit Performance 234 10.3 Fits to Blinded Dataset 237 10.4 Fits to Unblinded Dataset 238 10.5 Cross Section Limits for Pair Production of Vector-Like Top Quark 244 10.5.1 Overview of 0-Lepton Channels 247 10.5.2 Combination of 1-Lepton and 0-Lepton Channels 248 10.6 Cross Section Limits for Four-Top Production 252 10.7 Appendix 255 10.7.1 Plots in the Fit to the Asimov Dataset 255 10.7.2 Fit Stability Test 256 10.7.3 Impact on μ for Various Signal Models 257 10.7.4 Plots in the Fit to the Blinded Data 259 10.7.5 Plots in the Fit to the Partially Unblinded Data 260 10.7.6 Plots in the Fit to the Unblinded Data 262 Reference 266 11 Discussion 267 11.1 Categories Causing the Discrepancy 267 11.2 Validity of the Asymptotic Approximation 269 11.3 TRF Closure 270 11.4 Fit with Combined Categories 272 11.5 Correlation Scheme of the Systematic Uncertainties 273 11.6 Post-fit Kinematic Variables at the Preselection Level 274 11.7 Reweighing Leading Jet pT 276 11.8 Summary 279 Reference 279 12 Prospect 280 12.1 Sensitivity for High Luminosity 280 12.2 Performances of Higgs Boson and Top Quark Taggers 282 References 285 13 Conclusion 286 References 288 This book reports on the search for a new heavy particle, the Vector-Like Top quark (VLT), in the Large Hadron Collider (LHC) at CERN. The signal process is the pair production of VLT decaying into a Higgs boson and top quark (TT→Ht X, X=Ht, Wb, Zt). The signal events result in top–antitop quarks final states with additional heavy flavour jets. The book summarises the analysis of the data collected with the ATLAS detector in 2015 and 2016. In order to better differentiate between signals and backgrounds, exclusive taggers of top quark and Higgs boson were developed and optimised for VLT signals. These efforts improved the sensitivity by roughly 30%, compared to the previous analysis. The analysis outcomes yield the strongest constraints on parameter space in various BSM theoretical models. In addition, the book addresses detector operation and the evaluation of tracking performance. These efforts are essential to properly collecting dense events and improving the accuracy of the reconstructed objects that are used for particle identification. As such, they represent a valuable contribution to data analysis in extremely dense environments. Front Matter ....Pages i-xviii Introduction (Daiki Yamaguchi)....Pages 1-24 The ATLAS Experiment at the Large Hadron Collider (Daiki Yamaguchi)....Pages 25-51 Reconstruction of Physics Objects (Daiki Yamaguchi)....Pages 53-77 Data and Event Simulations (Daiki Yamaguchi)....Pages 79-83 Signal and Background Modeling (Daiki Yamaguchi)....Pages 85-131 Optimisations of Heavy Object Tagging and Event Categorisation (Daiki Yamaguchi)....Pages 133-174 Analysis Strategy (Daiki Yamaguchi)....Pages 175-197 Systematic Uncertainties (Daiki Yamaguchi)....Pages 199-208 Statistical Analysis (Daiki Yamaguchi)....Pages 209-215 Results (Daiki Yamaguchi)....Pages 217-255 Discussion (Daiki Yamaguchi)....Pages 257-269 Prospect (Daiki Yamaguchi)....Pages 271-276 Conclusion (Daiki Yamaguchi)....Pages 277-279