Differential top quark cross section results from the ATLAS and CMS experiments

This report summarizes recent results of differential top quark cross section measurements performed by the ATLAS and CMS experiments. The $t\bar{t}$ process is studied as well as the production of single (anti-)top quarks and the interference with other Standard Model processes of the same final state. State-of-the-art theory predictions are compared to the data. No theory model is able to describe the data across all bins, but an improved description of the data when moving to predictions in higher orders in perturbative QCD can be observed.

💡 Research Summary

This report compiles the most recent differential top‑quark cross‑section measurements performed by the ATLAS and CMS collaborations using the full Run 2 dataset at √s = 13 TeV (approximately 140 fb⁻¹ for ATLAS and 138 fb⁻¹ for CMS). The analyses cover top‑pair (t t̄) production in the single‑lepton (ℓ+jets) and dilepton (eμ) channels, jet‑substructure studies of boosted top jets, inclusive W b W b final‑state measurements that probe the interference between t t̄ and tW production, and single‑top production in the t‑channel for both top and anti‑top quarks.

In the ℓ+jets channel, CMS combines resolved reconstruction (pT(t) < 500 GeV) with a boosted topology (large‑R jets, R = 0.8, b‑tagged) and employs neural‑network classifiers to separate signal from background. A simultaneous fit across 18 orthogonal categories yields particle‑level and parton‑level differential cross sections. The results are generally well described by NLO+PS generators such as Powheg+Pythia/Herwig and MG5_aMC@NLO+Pythia, while NNLO QCD calculations provide an improved description, especially at high pT and high invariant mass, though residual discrepancies remain in the extreme tails.

The dilepton analysis uses a full kinematic reconstruction of the two neutrinos, regularized matrix unfolding (TUnfold), and provides single‑, double‑ and triple‑differential measurements as functions of lepton kinematics, the t t̄ system’s transverse momentum, invariant mass, and rapidity. The triple‑differential results constitute the first measurement of this dimensionality and are particularly sensitive to higher‑order QCD effects. NLO generators show sizable deviations in several observables, whereas NNLO and NNLO+NNLL predictions achieve a much better overall agreement.

ATLAS’s jet‑substructure study investigates eight observables of large‑R (R = 1.0) top jets with pT > 350 GeV, including Les Houches angularities and N‑jettiness. While angularity‑type variables are reproduced by most NLO+PS simulations, observables probing the three‑prong nature of top decays (e.g., N‑jettiness) reveal tensions, indicating the need for refined parton‑shower modeling and higher‑order corrections.

The inclusive W b W b measurement defines the final state directly as W b W b, thereby encompassing both t t̄ and tW contributions. Two interference modeling schemes—diagram removal (DR) and diagram subtraction (DS)—are tested. The m_bl_minimax observable maximizes sensitivity to interference effects. In the high‑mass region, DR overestimates the data while DS underestimates it; the dedicated bb4ℓ generator performs better but still falls short in the extreme tail. This analysis highlights the challenges of accurately simulating the full W b W b final state and the importance of improved interference modeling.

Single‑top t‑channel production is measured differentially in pT and absolute rapidity for both top and anti‑top quarks, together with the top/anti‑top cross‑section ratio. Neural‑network discriminants are used to enhance signal purity. The unfolded parton‑level results agree with fixed‑order calculations up to NNLO and with a variety of PDF sets, though systematic uncertainties are dominated by signal modeling.

Finally, ATLAS presents differential lepton distributions in t t̄ dilepton events, comparing several generators. State‑of‑the‑art Powheg MiNNLO and Powheg bb4ℓ provide a superior description relative to the older Powheg hvq model.

Across all measurements, theoretical predictions at NNLO QCD (and NNLO+NNLL where available) markedly improve agreement with data compared with NLO‑only calculations. Nevertheless, no single generator reproduces the full phase space; discrepancies persist in extreme kinematic regions and in higher‑dimensional observables. The results therefore provide valuable constraints for future Monte‑Carlo tuning, PDF fits, and for the development of even higher‑order calculations (e.g., N³LO, resummed approaches). Anticipated Run 3 data, with larger statistics and refined detector calibrations, are expected to further reduce experimental uncertainties and to enable more stringent tests of the Standard Model in the top‑quark sector.