Project Topics
Get Complete Project Material File(s) Now! »
Performance of the muon spectrometer
The pT resolution in the muon spectrometer was measured using cosmic ray data taken in 2009 by splitting tracks crossing the centre of the detector into two “collision-like” tracks and comparing their measured pT ’s. An example of the resolution is shown in Figure 2.22 obtained from one reconstruction algorithm in the sectors of the MDT located on the coils of the toroid. The data points are fitted to an empirical parameterization including energy loss fluctuation, multiple scattering and spectrometer resolution. The results obtained reach the designed goals.
Trigger and event selection
During the ATLAS startup phase, the focus of the trigger selection strategy was to commission the trigger and to ensure that well known Standard Model processes are observed. With peak luminosity less than few times of 1027 cm−2s−1, the minimum bias L1 trigger operated using hits in scintillator counters (MBTS), and HLT operated in pass-through mode. At peak luminosity around 1029 cm−2s−1, HLT chains have been activated to cope with increasing rate while running with low L1 thresholds. The trigger rates were kept about constant while the luminosity quickly increased by prescaling low threshold trigger items. An example of the L1 calorimetric trigger efficiency is shown in Figure 2.23. Good agreement between the collision data and MC-simulated sample is observed.
Material mapping in the ID
A very precise mapping of the material in the ID was obtained from the rate and vertex position of secondary hadronic interactions [70] or photon conversions [71] and from the energy flow [72] in the calorimeter. The three methods give results compatible with each other and show some discrepancy from the MC description of the detector. The maps obtained from data are used to improve the simulation.
The distribution of photon conversion vertices can be used to map the distribution of material in the ID. After applying some geometric selection criteria and the requirement on the fit quality of the conversion vertex, the position of selected photon conversion vertices are displayed in Figure 2.24. In this Figure, the beam pipe (R = 34.3 mm), the three barrel Pixel layers (R = 50.5, 88.5, 122.5 mm) and the first two SCT barrel layers (R = 299, 371 mm), together with the Pixel Support Tube (R = 229 mm) and various other support structures are clearly seen. In the xy projection, the cooling pipes on the Pixel detector modules and the overlap regions in the first SCT layer are visible. As shown in Figure 2.24 (b) clear shift in the simulated radial positions is observed for the Pixel Support Tube and global Pixel Support structure (around R = 200 mm), while the overall amount of material seems to be in good agreement.
Photon triggers in ATLAS
In ATLAS, electrons and photons are reconstructed by the trigger system in the range |η| <2.5. The selection variables for photons at each of the trigger levels are summarized in the following [78].
• level-1 (L1) selection
The whole spatial space (|η| < 2.5 and |ϕ| < π) is segmented into trigger towers. Each trigger tower has a size of Δη×Δϕ = 0.1×0.1. All the cells’ energy within the trigger tower are summed over all the layers of the electromagnetic and hadronic calorimeter. If the energy of four central trigger towers in a sliding window of 4 × 4 trigger towers is the local maximum with respect to its nearest overlapping neighbours, this 4 × 4 window is considered as a candidate. In the 2 × 2 tower core of this window, there are four combinations of two neighboring towers. If the energy in at least one of the combinations passes the electromagnetic cluster threshold (the threshold depends on the trigger item. e.g. for L1 EM5, it is 5 GeV ), it is then considered to contain an electron or photon candidate. Figure 4.1 shows the L1 trigger tower scheme. Isolation requirements can also be defined based on this scheme. The requirements to be imposed on the isolation energy are as follows:
{ The total transverse energy in the 12 towers of the electromagnetic calorimeter surrounding the 2 × 2 tower core is less than the electromagnetic isolation threshold.
{ The total transverse energy in the 4 towers of the hadronic calorimeter behind the 2 × 2 tower core of the electromagnetic calorimeter is less than the hadronic core threshold.
{ The total transverse energy in the 12 towers surrounding the 2×2 tower core of the hadronic calorimeter is less than the hadronic isolation threshold.
Table of contents :
1 Introduction
1.1 The Standard Model of particle physics
1.1.1 Gauge invariance in Quantum Electrodynamics
1.1.2 Electroweak theory
1.1.3 Introduction to QCD
1.1.4 The Higgs mechanism
1.1.5 Theoretical constraints on the Higgs mass
1.2 Review of Higgs searches
1.2.1 Direct searches
1.2.2 Indirect constraints
1.3 Review of di-photon cross-section measurements
1.4 Motivation and structure for this thesis
2 Accelerator and detector
2.1 Introduction to LHC
2.2 The ATLAS Detector
2.2.1 Coordinate system and nomenclature
2.2.2 Physics requirement and performance goals
2.2.3 Overview of the ATLAS detector
2.2.4 Magnet system
2.2.5 Inner detector
2.2.6 Calorimeter
2.2.7 Muon spectrometer
2.2.8 Trigger and data acquisition system
2.3 ATLAS data taking status and performance
2.3.1 Performance of the Inner Detector
2.3.2 Performance of the calorimeter
2.3.3 Performance of the muon spectrometer
2.3.4 Trigger and event selection
2.3.5 Material mapping in the ID
3 Photon reconstruction, calibration and identication
3.1 Photon reconstruction
3.2 Photon calibration
3.3 Photon identification
4 Photon Trigger Eciency
4.1 Introduction
4.2 Photon triggers in ATLAS
4.3 Triggers items and the corresponding selection criteria
4.4 Photon trigger efficiency measurement: description of the methods
4.4.1 The tag & probe method
4.4.2 The bootstrap method
4.4.3 The electron-to-photon extrapolation method
4.5 Monte Carlo study
4.5.1 Reconstructed photon selection criteria
4.5.2 Monte Carlo samples
4.5.3 Results of the tag&probe method
4.5.4 Tag&probe systematics
4.5.5 Results of bootstrap method
4.5.6 Bootstrap systematics
4.5.7 Comparison between the tag&probe and bootstrap samples
4.5.8 The electron to photon extrapolation method
4.5.9 Electron extrapolation systematics
4.5.10 Comparison of the three data-driven methods
4.5.11 Conclusion
4.6 Efficiency measurement on real data
4.6.1 g20 loose efficiency measurement
4.6.2 g10 loose efficiency for inclusive photon cross-section measurement .
5 Isolated di-photon cross-section measurement
5.1 Di-photon production and background processes
5.1.1 signal processes
5.1.2 Background process
5.2 Data and Monte Carlo samples
5.3 Photon and event selection
5.3.1 Photon isolation
5.3.2 Other photon identification criteria
5.3.3 Event selection
5.4 Extraction of the di-photon signal
5.4.1 Extraction of the yields
5.4.2 Signal isolation transverse energy one-dimensional PDF
5.4.3 Background isolation transverse energy one-dimensional PDF
5.4.4 Two-dimensional jj isolation transverse energy PDF
5.4.5 Tests on Monte Carlo
5.4.6 Results on real data
5.4.7 Systematic uncertainties
5.4.8 Differential spectra
5.5 Other methods
5.5.1 2×2D sideband
5.5.2 4×4 matrix
5.5.3 Comparison of the three methods
5.6 Differential cross-section
5.6.1 Extract the background from electrons
5.6.2 Trigger efficiency
5.6.3 Identification efficiency
5.6.4 Unfolding matrix
5.6.5 Reconstruction efficiency
5.6.6 Final result
6 H → γγ analysis
6.1 Signal and background processes
6.1.1 Higgs production
6.1.2 Higgs decay
6.1.3 Background processes
6.2 Discriminating variables
6.3 Principles of an early H → γγ analysis
6.3.1 Photon selection
6.3.2 Event selection
6.3.3 Primary vertex reconstruction
6.3.4 Extraction of the exclusion limit
6.4 Assessment of the exclusion limit from Monte Carlo simulation
6.4.1 Event-level trigger efficiency measurement
6.4.2 Signal and background estimation
6.4.3 Extrapolation to 7 TeV
6.4.4 Incorporating systematics uncertainties
6.5 Sensitivity on 2010 data
6.5.1 Trigger efficiency measurement
6.5.2 Analysis on real data: background decomposition using the di-photon analysis technique
6.5.3 Comparison with the Monte Carlo prediction
6.5.4 Projected sensitivity
6.6 Conclusion and prospects
