Measurement of the mass of the Higgs boson in the H→ZZ*→4l channel at the ATLAS detector

Presented in this thesis are the results from a measurement of the mass of the Higgs Boson in the H→ZZ*→4l (l = e, μ) channel using ATLAS data from pp collisions at a centre of mass energy of 13 TeV corresponding to an integrated luminosity of 139 fb−1. This measurement uses a Double-sided Crystal Ball function as an analytic signal model, incorporating per-event uncertainties as estimated using a quantile regression neural network. The mass of the Higgs boson is determined to be 124.92 ± 0.19 (stat.) +0.09 (syst.) GeV, in line with the previous LHC Run 1 combined result from the ATLAS and CMS collaborations of 125.09 ± 0.21 (stat.) ± 0.11 (syst.) GeV and the previous ATLAS measurement, which used 36 fb−1 of 13 TeV pp collision data, of 124.97 ± 0.18 (stat.) ± 0.20 (syst.) GeV.

Also presented are results from a method to validate the ATLAS photon energy calibration procedure by calibrating electrons from Z → ee events as photons. This technique is also used to calibrate unconverted photons as converted and vice versa to investigate the effect of misidentifying photon conversions. From this it is found that misidentifying all unconverted photons as converted photons which leave either one or two tracks in the transition radiation tracker (single/double TRT converted photons) leads to an η-independent bias in calibrated energy of approximately 1%. It is also found that misidentifying all single/double TRT converted photons as unconverted leads to a smaller bias, which is around 1% in the pseudorapidity region 0.8 ≤ |η| < 1.37.

Results from the application of shift and width adjustments to simulated electron shower shapes to correct for mismodelling are presented and a new method for per- forming such corrections using a Gaussian convolution technique is demonstrated. This new method is found to reduce the amount of over-correction in the application of corrections to electron shower shapes.