Measurement of the inclusive and differential cross section of a top quark pair in association with a Z boson at 13\,\text{TeV} with the ATLAS detector
aa r X i v : . [ h e p - e x ] J a n ATL-PHYS-PROC-2021-004January 27, 2021
Measurement of the inclusive and differential cross section ofa top quark pair in association with a Z boson at TeVwith the ATLAS detector
Florian Fischer , on behalf of the ATLAS Collaboration Fakultät für PhysikLudwig-Maximilians-Universität München, 85748 Garching, Germany
The inclusive as well as differential cross section of the associated pro-duction of top-antitop quark pairs and a Z boson ( ttZ ) is measured in fi-nal states with exactly three or four isolated leptons (electrons or muons).For this purpose, the full LHC Run 2 dataset of proton-proton collisionsrecorded by the ATLAS detector from to , which correspondsto an integrated luminosity of
139 fb − , is used. The inclusive productioncross section is measured to be σ ttZ = 1 . ± .
05 ( stat. ) ± .
09 ( syst. ) pb,which is in agreement with the most precise Standard Model theoreticalprediction. Absolute and normalised differential cross section measure-ments are performed as a function of various kinematic variables in or-der to probe the kinematics of the ttZ system within both parton- andparticle-level phase spaces. PRESENTED AT th International Workshop on Top Quark PhysicsDurham, UK (videoconference), 14–18 September, 2020 Work supported by BMBF, Germany (FSP-103)Copyright 2021 CERN for the benefit of the ATLAS Collaboration. CC-BY-4.0 license.
Introduction
The coupling of the top quark to the Z boson is precisely predicted within the Stan-dard Model (SM) of particle physics by the theory of the electroweak interaction.However, experimentally it is not yet well constrained and its value can significantlyvary in many models including physics beyond the Standard Model (BSM). A processthat is particularly sensitive to this coupling is the associated production of a top-antitop quark pair with a Z boson ( ttZ ). The large centre-of-mass energy of the LargeHadron Collider (LHC) [1] at CERN and the tremendous amount of data collectedin recent years have opened up the possibility to study this rare process which waspreviously inaccessible due to its small production cross section. As ttZ productioncontributes to the background processes in many searches at the LHC for both SMand BSM physics, a better understanding of the ttZ process can further enhance theexperimental reach in such analyses.The results of previous inclusive measurements by the ATLAS [2] and CMS [3]collaborations agree very well with the SM prediction [4, 5, 6]. A first measurementof differential ttZ cross sections was conducted by CMS only recently [7]. The firstanalysis using the full LHC Run 2 dataset was performed by ATLAS using
139 fb − of proton-proton ( pp ) collision data [8] which is presented in the following. The most sensitive decay channels in which to perform measurements of the ttZ process feature a multi-lepton final state with exactly three or four isolated electronsor muons.Based on these signatures, different signal regions are defined and optimised,referred to as trilepton ( ℓ ) and tetralepton ( ℓ ) signal regions, depending on therespective lepton multiplicity.Three signal regions are defined for the trilepton decay channel, and four signalregions are defined for the tetralepton decay channel. Of all lepton pairs with oppositesign of the charge and of the same flavour (OSSF), the one with the value of itsinvariant mass closest to the Z boson mass is considered to originate from the Z boson decay. Furthermore, the difference between its invariant mass and the Z bosonmass must not be greater than
10 GeV . Contributions from events featuring low-mass resonances are suppressed by requiring all OSSF lepton combinations to havea mass greater than
10 GeV . Additionally, the sum of the lepton charges is requiredto equal to ± and in the ℓ and in the ℓ case, respectively. The trilepton signalregions differ from each other by the number of selected jets and b -jets, where thelatter are tagged with different efficiency working points depending on the required b -jet multiplicity. Similarly, the tetralepton signal regions are categorised into same-flavour and different-flavour regimes of the two non- Z leptons, and each case is again1ubdivided into a regime with either exactly one or at least two b -jets. In addition,depending on the flavour composition of the non- Z lepton pair and b -jet multiplicity,different thresholds on the missing transverse energy are required. Background processes – physics processes described by the Standard Model otherthan ttZ – are subdivided into prompt and non-prompt contributions.The dominant prompt background processes are
W Z/ZZ + jets production whichfeature either three or four isolated leptons in the final state, respectively. Dedicatedcontrol regions are used to estimate the light-flavour components of these backgroundsduring the fit employed for the inclusive cross-section measurement. These regionsare defined such that they are orthogonal to the respective signal regions and arepredominantly populated with events featuring
W Z/ZZ + jets light-flavour compo-nents. In contrast, the charm- and bottom-flavour components are constrained in thefit with the corresponding uncertainties assigned which are related to the simulationof heavy-flavour components. Further SM background processes considered such asthe associated production of single top quarks or top-antitop quark pairs with heavyvector bosons are estimated directly from simulated Monte Carlo (MC) samples.Background contributions from leptons from secondary decays (“non-prompt”) orso-called fake leptons (objects misidentified as leptons), however, are estimated em-ploying a data-driven method, referred to as matrix method. Details about thismethod can be found in the reference documents [9] and [10].
The inclusive ttZ production cross section is extracted by performing a simultaneousmaximum-likelihood fit to the number of events in the trilepton and tetralepton signalregions, as well as the
W Z/ZZ + jets control regions. A total of three free param-eters are given to the fit: the ratio between the measured value of in the inclusive ttZ production cross section and its corresponding Standard Model prediction, re-ferred to as signal strength, as well as the normalisation factors of the
W Z/ZZ + jetsbackgrounds used to extrapolate the corresponding event yields into the signal re-gions. The inclusive ℓ + 4 ℓ cross section of ttZ production in pp -collision data at acentre-of-mass energy of
13 TeV is measured to be: σ ( pp → ttZ ) = 1 . ± .
05 ( stat. ) ± .
09 ( syst. ) pb = (1 . ± . pb (1)This result agrees with the dedicated theoretical prediction [11] of σ NLO+NNLL ttZ = 0 . +0 . − . (scale) ± .
03 (PDF + α s ) pb . (2)2he uncertainties on this result are dominated by the systematic uncertainties ofwhich the most important ones are related to the modelling of the parton showerin the signal Monte Carlo, the modelling of various background processes, and the b -tagging procedure.In addition to the inclusive result, the ttZ cross section is measured as a functionof different variables sensitive to the kinematics and the production of the ttZ system.For this purpose, a total of nine such variables are unfolded to parton and particlelevel, employing the Iterative Bayesian Unfolding method [12]. On parton level, the(anti-)top quark and the Z boson can be directly accessed before the decay withinMonte Carlo simulation, whereas on particle level these have to be reconstructedfrom simulated stable particles without any modelling of their interaction with thedetector material or pile-up. In this way, events are corrected for detector effectsand results can be directly compared to theoretical calculations. This analysis deter- d σ d p Z T [f b · G e V − ]
3l + 4l combination √ s = 13 TeV, 139 fb − ATLAS
Preliminary
DataMG5 aMc@NLO + Pythia8MG5 aMc@NLO + Herwig7Sherpa NLO inclusiveSherpa NLO multi-legNLO + NNLL JHEP 08 (2019) 039 p Z T [GeV]0.51.01.52.0 T heo r y D a t a Stat. Stat. ⊕ Syst. (a) σ · d σ d | y Z |
3l + 4l combination √ s = 13 TeV, 139 fb − ATLAS
Preliminary
DataMG5 aMc@NLO + Pythia8MG5 aMc@NLO + Herwig7Sherpa NLO inclusiveSherpa NLO multi-leg | y Z | T heo r y D a t a Stat. Stat. ⊕ Syst. (b)
Figure 1: Absolute (left) and normalised (right) cross section measured at parton(left) and particle (right) level as a function of the transverse momentum (left) and ofthe absolute rapidity (right) of the Z boson. The predictions of various MC generatorsare represented by dashed and dotted coloured lines whereas the data are depicted asblack dots. In addition, custom differential [15] predictions are shown by a black solidline within a grey-shaded area. In the ratio panels, the relative contributions fromboth the statistical and systematic uncertainties are shown. A branching fraction of B ( ttZ ℓ +4 ℓ ) = 0 . is applied for the parton-level result [8].mined both the absolute and normalised differential cross section for the ℓ and ℓ scenarios separately as well as for the combination. In Figure 1, two examples for the3ifferential cross section measurements in the combined ℓ + 4 ℓ channel are depicted.Different simulated samples generated with different sets of MC generators, includingthe nominal MG5_aMC@NLO+Pythia 8 [13, 14], as well as a set of additionaldifferential predictions, calculated at parton level as described in [15], are comparedto the unfolded data. In general, a good agreement between the unfolded data andthe various predictions can be observed.
ACKNOWLEDGEMENTS
The author would like to thank for the support of his work by BMBF, Germany(FSP-103).
References [1] L. Evans and P. Bryant (editors), JINST , S08001 (2008).[2] ATLAS Collaboration, JINST , S08003 (2008).[3] CMS Collaboration, JINST , S08004 (2008).[4] ATLAS Collaboration, Eur. Phys. J. C , no. 1, 40 (2017).[5] ATLAS Collaboration, Phys. Rev. D , no. 7, 072009 (2019).[6] CMS Collaboration, JHEP , 011 (2018).[7] CMS Collaboration, JHEP , 056 (2020).[8] ATLAS Collaboration, ATLAS-CONF-2020-028,https://cds.cern.ch/record/2725734.[9] ATLAS Collaboration, Eur. Phys. J. C , 1577 (2011).[10] ATLAS Collaboration, JHEP , 035 (2014).[11] A. Kulesza et al., Eur. Phys. J. C , no. 3, 249 (2019)[12] G. D’Agostini, Nucl. Instrum. Meth. A , 487 (1995).[13] J. Alwall et al., JHEP , 079 (2014)[14] T. Sjöstrand et al., Comput. Phys. Commun. , 159 (2015)[15] A. Broggio et al., JHEP1908