New Physics backgrounds to the H -> WW search at the LHC?

TL;DR

The study investigates whether new physics can bias data-driven background estimates for H → WW searches at the LHC by examining MSSM scenarios with light charginos and sleptons that populate the control region. Using event generation with Herwig++ and POWHEG, it simulates how SUSY processes mimic WW signatures and how they influence the control-to-signal extrapolation. The results show that BSM contributions can inflate the control-region background and distort the derived WW prediction in the signal region, especially in the 0-jet category and in the high m_T tail, with quantified factors indicating potential underestimation of the Higgs signal if not accounted for. The findings highlight the need to consider hard-region new physics when interpreting Higgs measurements and their couplings, since data-driven background estimates may be susceptible to contamination from beyond-SM processes.

Abstract

The searches for H -> WW events at the LHC use data driven techniques for estimating the q qbar -> WW background, by normalizing the background cross section to data in a control region. We investigate the possibility that new physics sources which mainly contribute to the control region lead to an overestimate of Standard Model backgrounds to the Higgs boson signal and, thus, to an underestimate of the H -> WW signal. A supersymmetric scenario with heavy squarks and gluinos but charginos in the 200 to 300 GeV region and somewhat lighter sleptons can lead to such a situation.

Figures (3)

• Figure 1: Invariant lepton pair mass $m_{\ell\ell}$ (left) and transverse mass $m_T$ (right) distributions for the "base" scenario of Section \ref{['scenario']}. The $m_T$ distribution is calculated including all control region cuts of Eqs. \ref{['basecuts']} - \ref{['etmissrelcut']}, \ref{['ptllcut']} and \ref{['mllcut']}. For the $m_{\ell\ell}$ plot the $m_{\ell\ell}$-cut of Eq. \ref{['mllcut']} is omitted. Both plots show the $q\bar{q}\rightarrow WW$ distribution, the SUSY contributions and their sum. The $m_{\ell\ell}$ plot also shows the $q\bar{q}\rightarrow WW$ result, rescaled by $(\sigma_C^{WW}+\sigma_C^{SUSY}) / \sigma_C^{WW}$, extracted from the control region.
• Figure 2: Event numbers of the SUSY contributions in the signal region, in the control region and in the control region with $m_T > 350\,\textrm{GeV}$ and $m_T > 440\,\textrm{GeV}$ for varying slepton masses of the first two generations. The LSP mass is $m_{\chi_1^0}=99\,\textrm{GeV}$ in the left plot and $m_{\chi_1^0}=124\,\textrm{GeV}$ in the right plot. The discussed "base" and "worst case" scenarios are marked.
• Figure 3: Invariant lepton pair mass $m_{\ell\ell}$ (left) and transverse mass $m_T$ (right) distributions for the "worst case" scenario of Section \ref{['results']}. The $m_T$ distribution is calculated including all control region cuts of Eqs. \ref{['basecuts']} - \ref{['etmissrelcut']}, \ref{['ptllcut']} and \ref{['mllcut']}. For the $m_{\ell\ell}$ plot the $m_{\ell\ell}$-cut of Eq. \ref{['mllcut']} is omitted. Both plots show the $q\bar{q}\rightarrow WW$ distribution, the SUSY contributions and their sum. The $m_{\ell\ell}$ plot also shows the $q\bar{q}\rightarrow WW$ result, rescaled by $(\sigma_C^{WW}+\sigma_C^{SUSY}) / \sigma_C^{WW}$, extracted from the control region.