First measurement of the associated production of a Z boson with b jets at the LHC

The start of the LHC in 2009 opened a new era in particle physics. The collisions, produced at an energy and a rate unequalled, allowing to probe energy sectors and rare processes so far inaccessible. In summer 2012 the announcement of the discovery a new boson, presenting the properties of the long-sought Higgs boson was confirming the last prediction of the particle physics theoretical framework called the standard model. This theory, even though highly predictive and never defeated experimentally, is known to have limitations and lacks explanation for several physics observations. The production of a Z boson in association with b jets is a process for which different theoretical predictions from the standard model differ, depending on the treatment of the b jets in the calculation (so-called 4- and 5-flavour schemes). The measurement of the Z plus b jets cross section and the study of its kinematics constitute then a strong QCD test. The Z plus at least two b jets process constitutes the main background for studies of the newly discovered boson, as well as for searches in theories beyond the standard model, presenting the same final state. In this thesis the cross section measurement is done both for Z plus exactly one b jet and Z plus at least two b jets, together giving the result for Z plus at least one b jet. The data used, corresponding to an integrated luminosity of 5 fb−1 , were recorded by the CMS detector from proton-proton collisions at a centre-of-mass energy of 7 TeV at the LHC. The event selection is led by the care to optimise the signal over background ratio and to select robust well identified objects. b-tagging techniques are used for the selection of the b jets. A special attention is given to the background estimation, partially data-driven when possible. To yield a cross section at the particle level, all the efficiencies of reconstruction and selection and the detector effects are estimated taking into account the migrations in terms of b jet multiplicity. The results are compared with several theoretical predictions coming from calculations at leading order or next-to-leading order and in the two schemes available. Important kinematic variables shown at the reconstructed level are also deconvoluted at the generator level. For this, two different unfolding techniques are explained and used: the Singular Value Decomposition and the Bayesian unfolding. The analysis selection was implemented in the RIVET framework, so that the differential cross sections can be compared to any generator.