Studies of jet mass in dijet and W/Z + jet events

TL;DR

This CMS study investigates jet mass distributions in dijet and V+jet events at 7 TeV, applying grooming techniques (filtering, trimming, pruning) to anti-kt and Cambridge-Aachen clustered jets. Through unfolding to particle level and comparisons to LO QCD with various parton-shower MCs (Pythia6/8, Herwig++ , MadGraph), the paper demonstrates that grooming reduces sensitivity to pileup and improves data–MC agreement, with pruning typically providing the strongest improvement. The results show that V+jet events (dominated by quark jets) are described slightly better by MC than dijet events (gluon-dominated), and that more aggressive grooming enhances agreement across kinematic ranges, particularly in the low-mass region. These measurements, including the first use of trimmed and pruned jets at the LHC, provide essential benchmarks for boosted-object searches and for validating MC models in high-luminosity environments.

Abstract

Invariant mass spectra for jets reconstructed using the anti-kt and Cambridge-Aachen algorithms are studied for different jet "grooming" techniques in data corresponding to an integrated luminosity of 5 inverse femtobarns, recorded with the CMS detector in proton-proton collisions at the LHC at a center-of-mass energy of 7 TeV. Leading-order QCD predictions for inclusive dijet and W/Z+jet production combined with parton-shower Monte Carlo models are found to agree overall with the data, and the agreement improves with the implementation of jet grooming methods used to distinguish merged jets of large transverse momentum from softer QCD gluon radiation.

Figures (23)

• Figure 1: Distributions in differential probability for ratios of the jet mass of groomed jets to their corresponding ungroomed values, for both dijet data and pythia6 (tune Z2) MC simulation, for the three grooming techniques discussed in the text: (i) filtering (circles, peaking near $0.9$), (ii) trimming (squares, peaking near $0.75$), and (iii) pruning (triangles, more dispersed).
• Figure 2: The $p_{\mathrm{T}}\xspace$ distribution for the leading AK7 jet in accepted (a) ${Z}$+jet and (b) $\mathrm{W}$+jet events.
• Figure 3: Distributions of the average jet mass for AK jets as a function of the number of reconstructed primary vertices: (a) for different jet radii, and (b) for AK7 jets, comparing the impact of grooming algorithms to results without grooming.
• Figure 4: Unfolded distributions for the mean mass of the two leading jets in dijet events for reconstructed AK7 jets, separated according to intervals in $p_{\mathrm{T}}\xspace^\mathrm{AVG}$ (the mean $p_{\mathrm{T}}\xspace$ of the two jets). The data are shown by the symbols indicating different bins in the mean $p_{\mathrm{T}}\xspace$ of the two jets. The statistical uncertainty is shown in light shading, and the total uncertainty in dark shading. Predictions from herwig++ are given by the dotted lines. To enhance visibility, the distributions for larger values of $p_{\mathrm{T}}\xspace^\mathrm{AVG}$ are scaled up by the factors given in the legend.
• Figure 5: Unfolded distributions for the mean mass of the two leading jets in dijet events for reconstructed filtered AK7 jets, separated according to intervals in $p_{\mathrm{T}}\xspace^\mathrm{AVG}$ (the mean $p_{\mathrm{T}}\xspace$ of the two jets). The data are shown by the symbols indicating different bins in the mean $p_{\mathrm{T}}\xspace$ of the two jets. The statistical uncertainty is shown in light shading, and the total uncertainty in dark shading. Predictions from herwig++ are given by the dotted lines. To enhance visibility, the distributions for larger values of $p_{\mathrm{T}}\xspace^\mathrm{AVG}$ are scaled up by the factors given in the legend.