Jet Substructure as a New Higgs-Search Channel at the Large Hadron Collider

It is widely considered that, for Higgs boson searches at the CERN Large Hadron Colider, $WH$ and $ZH$ production where the Higgs boson decays to $b\overline{b}$ are poor search channels due to large backgrounds. We show that at high transverse momenta, employing state-of-the-art jet reconstruction and decomposition techniques, these processes can be recovered as promising search channels for the standard model Higgs boson around 120 GeV in mass.

Figure 1

The three stages of our jet analysis: starting from a hard massive jet on angular scale $R$, one identifies the Higgs neighborhood within it by undoing the clustering (effectively shrinking the jet radius) until the jet splits into two subjets each with a significantly lower mass; within this region, one then further reduces the radius to $R_{\mathrm{filt}}$ and takes the three hardest subjets, so as to filter away UE contamination while retaining hard perturbative radiation from the Higgs decay products.

Figure 2

Signal and background for a 115 GeV SM Higgs signal simulated using HERWIG, CA MD-F with $R=1.2$ and $p_T>200\mathrm{GeV}$, for $30{\mathrm{fb}}^{-1}$. The $b$ tag efficiency is assumed to be 60%, and a mistag probability of 2% is used. The $q\overline{q}$ sample includes dijets and $t\overline{t}$. The vector boson selections for (a), (b), and (c) are described in the text, and (d) shows the sum of all three channels. The errors reflect the statistical uncertainty on the simulated samples and correspond to integrated luminosities $>30{\mathrm{fb}}^{-1}$.

Figure 3

Estimated sensitivity for $30{\mathrm{fb}}^{-1}$ under various different sets of cuts and assumptions (a) for $m_H=115\mathrm{GeV}$ as a function of the mistag probability for $b$-subjets and (b) as a function of Higgs mass for the $b$-tag efficiency (mistag rates) shown in the legend. Significance is estimated as $\mathrm{\text{signal}}/\sqrt{\mathrm{\text{background}}}$ in the peak region.