Transverse momentum resummation effects in W^+W^- measurements

TL;DR

The paper investigates approximate NNLL transverse-momentum resummation for W^+W^- production at the LHC and its impact on fiducial and inclusive cross sections. By reweighting NLO MC events with the resummed p_T distribution, the authors quantify a 3–7% increase in predicted cross sections, with the jet-veto cut emerging as the primary driver of this effect and the results depending on the generator and shower used. While the reweighting improves agreement between SM predictions and data for several MC setups, it does not fully account for the ~20% discrepancy observed, indicating a need for jet-veto resummation and further cross-channel studies. The work underscores the importance of accurate p_T shapes in diboson backgrounds and provides a framework and data for testing resummation effects in Ws and related channels.

Abstract

The W^+W^- cross section has remained one of the most consistently discrepant channels compared to SM predictions at the LHC, measured by both ATLAS and CMS at 7 and 8 TeV. Developing a better modeling of this channel is crucial to understanding properties of the Higgs and potential new physics. In this paper we investigate the effects of NNLL transverse momentum resummation in measuring the W^+W^- cross section. In the formalism we employ, transverse momentum resummation does not change the total inclusive cross section, but gives a more accurate prediction for the p_T distribution of the diboson system. By re-weighting the p_T distribution of events produced by Monte Carlo generators, we find a systematic shift that decreases the experimental discrepancy with the SM prediction by approximately 3-7% depending on the MC generator and parton shower used. The primary effect comes from the jet veto cut used by both experiments. We comment on the connections to jet veto resummation, and other methods the experiments can use to test this effect. We also discuss the correlation of resummation effects in this channel with other diboson channels. Ultimately p_T resummation improves the agreement between the SM and experimental measurements for most generators, but does not account for the measured ~20% difference with the SM and further investigations into this channel are needed.

Figures (9)

• Figure 1: Plot of resummed, finite (matching) and fixed-order $W^+W^-$ transverse momentum distributions from 8 TeV proton collisions.
• Figure 2: Plot of renormalization, factorization and resummation scale variations of the $W^+W^-$ transverse momentum distribution for 8 TeV collisions.
• Figure 3: NNLO+LO predictions, with error bands, for the $W^+W^-$ transverse momentum distribution for 7,8 and 14 TeV collisions.
• Figure 4: NNLL+LO prediction for the WW transverse momentum distribution at 8 TeV, with and without the non-perturbative Gaussian smearing factor $\exp \left[ -1\, {\rm GeV}^2\, b^2 \right]$.
• Figure 5: Comparision of our resummed WW $p_T$ distribution with a SCET-based resummation calculation, with error bands shown for both.