Bump-hunting in LHC ttbar events

TL;DR

This paper proposes a normalized NNLO top pair invariant mass spectrum as a precise tool for bump-hunting in LHC top quark events, capable of identifying or constraining new resonances like the 750 GeV di-gamma excess.

Contribution

It introduces a low-uncertainty, normalized NNLO spectrum for top pair mass, enhancing sensitivity to new resonances and distinguishing them from the Standard Model background.

Findings

High significance distinction between signal and background.

Potential to set strong limits on resonance decay rates.

Effective for future resonance searches in top quark events.

Abstract

We demonstrate that a purposefully normalised NNLO top pair invariant mass differential spectrum can have very small theoretical uncertainty and, in particular, a small sensitivity to the top quark mass. Such observable can thus be a very effective bump-hunting tool for resonances decaying to top pair events during LHC Run II and beyond. To illustrate how the approach works, we concentrate on one specific example of current interest, namely, the possible 750 GeV di-gamma excess resonance Phi. Considering only theoretical uncertainties, we demonstrate that it is possible to distinguish pp -> Phi -> tt signals studied in the recent literature [Hespel, Maltoni and Vryonidou, arXiv:1606.04149] from the pure SM background with very high significance. Alternatively, in case of non-observation, a strong upper limit on the decay rate Phi -> tt can be placed.