The process gg -> WW as a background to the Higgs signal at the LHC
M. D"uhrssen, K. Jakobs, P. Marquard, J. J. van der Bij
TL;DR
The paper tackles the challenging background from gluon fusion to $W$-pair production in the Higgs search channel $gg\to H\to WW$ at the LHC. It provides a full one-loop calculation of $gg\to WW$ with finite top-mass effects and preserves spin correlations in the leptonic decays, comparing this background to $q\bar{q}\to WW$ and the Higgs signal using PYTHIA. The study finds that the $gg$ background, though subleading after some cuts, remains sizable and is difficult to normalize experimentally, with heavy-quark effects being small. Consequently, the Higgs discovery significance in the $gg\to H\to WW$ mode is reduced, underscoring the need for precise theoretical control of the $gg\to WW$ background. The analysis employs a two-stage event selection (preselection and final cuts) to optimize signal-to-background in a LO framework.
Abstract
The production of W pairs from the one-loop gluon fusion process is studied. Formulas are presented for the helicity amplitudes keeping the top mass finite, but all other quark masses zero. The correlations among the leptons coming from the W bosons are kept. The contribution of this background to the Higgs boson search in the WW decay mode at the LHC is estimated by applying the cuts foreseen in experimental searches using the PYTHIA Monte Carlo program. Kinematic distributions for the final state leptons are compared to those of the Higgs boson signal and of the q qbar -> WW background. After applying final cuts, the gg background is found to be large, at the level of 35% of the q qbar background.The characteristics of the gg background are very similar to those of the signal. Therefore, an experimental normalization of this background component appears to be very difficult and the uncertainty must largely be determined by theory. As a result, the significance of a Higgs signal in the gg -> H -> WW mode at the LHC is reduced.