Soft-drop grooming for hadronic event shapes

TL;DR

This paper investigates soft-drop grooming of hadronic event shapes, demonstrating its effectiveness in reducing non-perturbative effects and enabling more accurate comparisons between theoretical predictions and experimental data.

Contribution

It derives NLL resummation formulas for soft-drop groomed event shapes and provides NLO+NLL' predictions using the Sherpa framework, advancing precision in QCD analyses.

Findings

Soft-drop grooming significantly reduces underlying event contributions.

Predictions at NLO+NLL' accuracy match well with parton- and hadron-level data.

Soft-drop parameters can be tuned to mitigate non-perturbative corrections.

Abstract

Soft-drop grooming of hadron-collision final states has the potential to significantly reduce the impact of non-perturbative corrections, and in particular the underlying-event contribution. This eventually will enable a more direct comparison of accurate perturbative predictions with experimental measurements. In this study we consider soft-drop groomed dijet event shapes. We derive general results needed to perform the resummation of suitable event-shape variables to next-to-leading logarithmic (NLL) accuracy matched to exact next-to-leading order (NLO) QCD matrix elements. We compile predictions for the transverse-thrust shape accurate to NLO+NLL' using the implementation of the CAESAR formalism in the Sherpa event generator framework. We complement this by state-of-the-art parton- and hadron-level predictions based on NLO QCD matrix elements matched with parton showers. We explore the potential to mitigate non-perturbative corrections for particle-level and track-based measurements of transverse thrust by considering a wide range of soft-drop parameters. We find that soft-drop grooming indeed is very efficient in removing the underlying event. This motivates future experimental measurements to be compared to precise QCD predictions and employed to constrain non-perturbative models in Monte-Carlo simulations.