On jet mass distributions in Z+jet and dijet processes at the LHC

TL;DR

This work advances jet-substructure phenomenology by delivering an analytic resummation of jet-mass distributions for Z+jet and dijet production at the LHC to NLL accuracy, including initial-state radiation and non-global logarithms, and matching to LO. The authors construct a dipole-based eikonal framework, derive explicit global radiators with $R$-dependent corrections, and implement a large-$N_c$ all-order treatment of non-global effects, complemented by LO matching and partial ${C_1}$-term estimates. They quantify the impact of ISR and non-global logs on the Sudakov peak, compare with several MC event generators, and examine jet-radius dependencies for phenomenology at $ oot s = 7$ TeV. The results offer improved perturbative control for jet-mass observables and establish a path toward NLO matching and broader substructure applications at the LHC.

Abstract

The mass distribution of jets produced in hard processes at the LHC plays an important role in several jet substructure related studies involving both Standard Model and BSM physics, especially in the context of boosted heavy particle searches. We compute analytically the jet-mass distribution for both Z+jet and dijet processes, for QCD jets defined in the anti-k_t algorithm with an arbitrary radius R, to next-to-leading logarithmic accuracy and match our resummed calculation to full leading-order results. We note the important role played by initial state radiation (ISR) and non-global logarithms explicitly computed here for the first time for hadron collider observables, as well as the jet radius dependence of these effects. We also compare our results to standard Monte Carlo event generators and discuss directions for further studies and phenomenology.

Figures (7)

• Figure 1: Comparison between different approximations to the resummed exponent: jet functions (blue), with full resummation of the global contribution (green) and with non-global logarithms as well (red). The jet radius is $R=0.6$.
• Figure 2: Comparison between different approximations to the resummed exponent: jet functions (blue), with full resummation of the global contribution (green) and with non-global logarithms as well (red). The jet radius is $R=1.0$.
• Figure 3: Matching of the NLL resummed distribution to the LO one for $R=0.6$ (on the left) and $R=1.0$ (on the right).
• Figure 4: The impact of the NLL constant $C_1$, for $R=0.6$ jets (on the left) and $R=1.0$ (on the right). The band is produced by suppressing the real radiation contributions with a quark or gluon jet form factor, as explained in the text.
• Figure 5: Comparison of our resummed and matched result NLL+LO (in red) to standard Monte Carlo event generators, at the parton level.