Improving the sensitivity of Higgs boson searches in the golden channel

TL;DR

The paper tackles the limitation of traditional Higgs searches in the golden channel by exploiting full angular information through the Matrix Element Method. It derives fully differential, Lorentz-invariant cross sections for both signal and irreducible ZZ background, including off-shell Z bosons, and implements an extended likelihood approach to combine kinematic shapes with event yields. Through a comprehensive Monte Carlo study at √s = 7 TeV for m_h in 175–350 GeV, it demonstrates that incorporating angular correlations yields about a 10–20% improvement in discovery significance and exclusion limits, particularly at higher Higgs masses. The work argues for adopting angular-distribution–aware multivariate analyses in experimental searches to enhance sensitivity in the golden channel.

Abstract

Leptonic decays of the Higgs boson in the ZZ* channel yield what is known as the golden channel due to its clean signature and good total invariant mass resolution. In addition, the full kinematic distribution of the decay products can be reconstructed, which, nonetheless, is not taken into account in traditional search strategy relying only on measurements of the total invariant mass. In this work we implement a type of multivariate analysis known as the matrix element method, which exploits differences in the full production and decay matrix elements between the Higgs boson and the dominant irreducible background from q bar{q} -> ZZ*. Analytic expressions of the differential distributions for both the signal and the background are also presented. We perform a study for the Large Hadron Collider at sqrt{s}=7 TeV for Higgs masses between 175 and 350 GeV. We find that, with an integrated luminosity of 2.5 fb^-1 or higher, improvements in the order of 10 - 20 % could be obtained for both discovery significance and exclusion limits in the high mass region, where the differences in the angular correlations between signal and background are most pronounced.

Figures (8)

• Figure 1: (a) Two decay planes of $Z_i\to \ell_i\bar{\ell}_i$, $i=1,2$. The polar angles $\theta_{i}$ shown are defined in the rest frames of $Z_{i}$ with respect to $\hat{k}_{i}$, while the azimuthal angles shown are in fact $2\pi-\phi_{1} = -\phi_{1}$ and $\pi-\phi_2$. (b) The coordinate system in the CM frame and the definition of the production angle $\Theta$.
• Figure 2: Feynman diagrams contributing to $q\bar{q}\to 4\ell$. We only consider (a), since final states from (b) have a total invariant mass at the $Z$ mass.
• Figure 3: Signal and background singly differential distributions at $m_h = \sqrt{{s}}= 220$ and $350$ GeV. The blue (dashed) lines are background distributions and the red (solid) lines are signal distributions.
• Figure 4: Doubly differential distributions for signal and background. See the text for more explanations.
• Figure 5: Effect of detector resolution in measuring $E$ and $p_T$ on kinematic variables. Each of the histograms were generated using twenty thousand events. The channel used for these plots is the background $2e2\mu$ channel.