Evidence for the associated production of the Higgs boson and a top quark pair with the ATLAS detector

A search for the associated production of the Higgs boson with a top quark pair ($t\overline{t}H$) is reported. The search is performed in multilepton final states using a data set corresponding to an integrated luminosity of $36.1{\mathrm{fb}}^{-1}$ of proton-proton collision data recorded by the ATLAS experiment at a center-of-mass energy $\sqrt{s}=13\mathrm{TeV}$ at the Large Hadron Collider. Higgs boson decays to $W{W}^{*}$, $\tau \tau$, and $Z{Z}^{*}$ are targeted. Seven final states, categorized by the number and flavor of charged-lepton candidates, are examined for the presence of the Standard Model Higgs boson with a mass of 125 GeV and a pair of top quarks. An excess of events over the expected background from Standard Model processes is found with an observed significance of 4.1 standard deviations, compared to an expectation of 2.8 standard deviations. The best fit for the $t\overline{t}H$ production cross section is $\sigma (t\overline{t}H)=790_{-210}^{+230}\mathrm{fb}$, in agreement with the Standard Model prediction of $507_{-50}^{+35}\mathrm{fb}$. The combination of this result with other $t\overline{t}H$ searches from the ATLAS experiment using the Higgs boson decay modes to $b\overline{b}$, $\gamma \gamma$ and $Z{Z}^{*}\to 4\ell$, has an observed significance of 4.2 standard deviations, compared to an expectation of 3.8 standard deviations. This provides evidence for the $t\overline{t}H$ production mode.

Figure 1

Examples of tree-level Feynman diagrams for the production of the Higgs boson in association with a pair of top quarks. Higgs boson decays to (a) $WW/ZZ$ or (b) $\tau \tau$ are shown.

Figure 2

The efficiency to select well-identified prompt muons (left) and electrons (right) at the chosen nonprompt lepton BDT working point, as a function of the lepton $p_{\mathrm{T}}$. The muons are required to pass the loose identification requirements, while the electrons are required to pass the tight identification requirements. The measurements in data (simulation) are shown as full black (open red) circles. The bottom panel displays the ratio of data to simulation results, with the blue (yellow) band representing the statistical (total) uncertainty. This ratio is the scale factor that is applied to correct the simulation.

Figure 3

The channels used in the analysis organized according to the number of selected light leptons and ${\tau }_{\mathrm{had}}$ candidates. The selection requirements for each channel are in Table 3.

Figure 4

Left: The fraction of the expected $t\overline{t}H$ signal arising from different Higgs boson decay modes in each signal region. The decays labeled as “other” are mostly $H\to \mu \mu$ and $H\to b\overline{b}$. Right: Prefit $S/B$ (black line) and $S/\sqrt{B}$ (red dashed line) ratios for each of the 12 analysis categories including the four $3\ell$ control regions. The background prediction methods are described in Sec. 6.

Figure 5

The fractional contributions of the various backgrounds to the total predicted background in each of the 12 analysis categories. The background prediction methods are described in Sec. 6: “Nonprompt,” “Fake ${\tau }_{\mathrm{had}}$ ” and “$q$ mis-id” refer to the data-driven background estimates (largely $t\overline{t}$ but also include other electroweak processes), and rare processes ($tZ$, $tW$, $tWZ$, $t\overline{t}WW$, triboson production, $t\overline{t}t$, $t\overline{t}t\overline{t}$, $tH$, rare top decay) are labeled as “Other.”

Figure 6

Comparison of data and prediction of the jet multiplicity in the (left) $3\ell t\overline{t}Z$ and the (right) $3\ell t\overline{t}W$ control regions. The last bin in each figure contains the overflow. The bottom panel displays the ratio of data to the total prediction. The hatched area represents the total uncertainty in the background. The background prediction for nonprompt leptons is described in Sec. 6b and the other backgrounds are normalized according to the predictions from simulation.

Figure 7

The composition from simulation of (a) the fake and nonprompt light leptons and (b) the fake ${\tau }_{\mathrm{had}}$ in selected analysis regions. The light-lepton composition is shown separately depending on the lepton flavor in the regions used in the estimate of the nonprompt contribution. The control regions labeled 2lSSxx are used for the $2\ell \mathrm{SS}$ and $3\ell$ channels; those labeled 3lx are used for the $4\ell$ channel, where x denotes the flavor of the lowest-$p_{\mathrm{T}}$ lepton, and those labeled $2\mathrm{lSSx}+1\tau$ are used for the $2\ell \mathrm{SS}+1{\tau }_{\mathrm{had}}$ channel. The nonprompt lepton background has been separated into the components from $b$ jets, $c$ jets, other jets, $J/\psi$, photon conversions and other contributions. The latter includes pion, kaon and nonprompt tau decays and cases where reconstructed leptons cannot be assigned unambiguously to a particular source. The ${\tau }_{\mathrm{had}}$ composition is shown both in the control regions used in the estimates and in the signal regions of each channel. The ${\tau }_{\mathrm{had}}$ background has been separated into the components from $b$ jets, $c$ jets, light-quark jets, gluon jets, electrons and other contributions. The latter includes muons, hadrons and cases where reconstructed leptons cannot be assigned unambiguously to a particular source.

Figure 8

Comparison of data and prediction of (a) the angular distance between the subleading lepton and the closest jet in the $\mu \mu$ channel and (b) the subleading lepton $p_{\mathrm{T}}$ in the opposite-flavor channel, in a $2\ell \mathrm{SS}$ low-$N_{\text{jets}}$ validation region (VR); (c) the $b$-tagged jet multiplicity in a validation region similar to the control region used in the $4\ell$ channel but at higher $N_{\text{jets}}$ multiplicity (called $3\ell \mathrm{VR}$), with the leptons categorized according to their origin: prompt, heavy-flavor (HF) and light-flavor (LF), see the text; (d) the jet multiplicity in the $2\ell \mathrm{OS}+1{\tau }_{\mathrm{had}}$ category. The last bin in each figure contains the overflow. The bottom panel displays the ratio of data to the total prediction. The hatched area represents the total uncertainty in the background.

Figure 9

The impact of systematic uncertainties on the fitted signal-strength parameter $\widehat{\mu }$ for the combined fit of all channels. The systematic uncertainties are listed in decreasing order of their impact on $\widehat{\mu }$ on the $y$ axis, and only the 15 most important ones are displayed. The filled blue boxes show the variations of $\widehat{\mu }$ from the central value, $\mathrm{\Delta }\mu$, referring to the upper $x$ axis, when fixing the corresponding individual nuisance parameter, $\theta$, to its postfit value $\widehat{\theta }$ modified upwards or downwards by its postfit uncertainty, and repeating the fit. The empty blue boxes represent the corresponding prefit impact. The black points, which refer to the lower $x$ axis, show the fitted values and uncertainties of the nuisance parameters, relative to their prefit values, ${\theta }_0$, and uncertainties, $\mathrm{\Delta }\theta$. The black lines show the postfit uncertainties of the nuisance parameters, relative to their nominal uncertainties, which are indicated by the dashed line.

Figure 10

Comparison of prediction to data after the fit in the eight signal and four control regions. The background contributions after the global fit are shown as filled histograms. The total background before the fit is shown as a dashed blue histogram. The Higgs boson signal ($m_H=125\mathrm{GeV}$), scaled according to the results of the fit, is shown as a filled red histogram superimposed on the fitted backgrounds. The size of the combined statistical and systematic uncertainty in the sum of the signal and fitted background is indicated by the blue hatched band. The ratio of the data to the sum of the signal and fitted background is shown in the lower panel. The yields in each region are shown in Table 10.

Figure 11

The distribution of the discriminating variables observed in data (points with bars indicating the statistical errors) and expected (histograms) in the (a) $2\ell \mathrm{SS}$, (b) $3\ell$, (c) $4\ell$ ($Z$-enriched) and (d) $4\ell$ ($Z$-depleted) signal regions. The background contributions after the global fit are shown as filled histograms. The total background before the fit is shown as a dashed blue histogram. The Higgs boson signal ($m_H=125\mathrm{GeV}$), scaled according to the results of the fit, is shown as a filled red histogram superimposed on the fitted backgrounds. The size of the combined statistical and systematic uncertainty in the sum of the signal and fitted background is indicated by the blue hatched band. The ratio of the data to the sum of the signal and fitted background is shown in the lower panel.

Figure 12

The distribution of the discriminating variables observed in data (points with bars indicating the statistical errors) and expected (histograms) in the (a) $2\ell \mathrm{SS}+1{\tau }_{\mathrm{had}}$, (b) $2\ell \mathrm{OS}+1{\tau }_{\mathrm{had}}$, (c) $1\ell +2{\tau }_{\mathrm{had}}$ and (d) $3\ell +1{\tau }_{\mathrm{had}}$ signal regions. The background contributions after the global fit are shown as filled histograms. The total background before the fit is shown as a dashed blue histogram. The Higgs boson signal ($m_H=125\mathrm{GeV}$), scaled according to the results of the fit, is shown as a filled red histogram superimposed on the fitted backgrounds. The size of the combined statistical and systematic uncertainty in the sum of the signal and fitted background is indicated by the blue hatched band. The ratio of the data to the sum of the signal and fitted background is shown in the lower panel.

Figure 13

The observed best-fit values of the $t\overline{t}H$ signal strength $\mu$ and their uncertainties by final-state category and combined. The individual $\mu$ values for the channels are obtained from a simultaneous fit with the signal-strength parameter for each channel floating independently. The SM prediction is $\mu =1$.

Figure 14

Event yields as a function of ${\log }_{10}(S/B)$ for data, background and a Higgs boson signal with $m_{\mathrm{H}}=125\mathrm{GeV}$. The discriminant bins in all signal regions are combined into bins of ${\log }_{10}(S/B)$, where $S$ is the expected signal yield and $B$ the background yield from the unconditional fit. The background yields are shown as the fitted values, while the signal yields are shown for the fitted value ($\mu =1.6$) and the SM prediction ($\mu =1$). The total background before the fit is shown as a dashed blue histogram. The pull (residual divided by its uncertainty) of the data relative to the background-only prediction is shown in the lower panel, where the full red line (dashed orange line) indicates the pull of the prediction for signal with $\mu =1.6$ ($\mu =1$) and background relative to the background-only prediction. The background is also shown after the fit to data assuming zero signal contribution as well as its pull (dotted black line) relative to the background from the nominal fit.

Figure 15

Summary of the measurements of $\mu$ from individual analyses and the combined result. “ML” refers to the multileptonic decay channels discussed in Sec. 8. The best-fit values of $\mu$ for the individual analyses are extracted independently, and systematic uncertainty nuisance parameters are only correlated for the combination. As no events are observed in the $H\to 4\ell$ analysis, a 68% confidence level (C.L.) upper limit on $\mu$, computed using the $\mathrm{C}.\mathrm{L}{.}_{\mathrm{s}}$ method [108], is reported.

Figure 16

Summary of the best-fit values of $\mu$ broken down by Higgs boson decay mode. The decays $H\to W{W}^{*}$ and $H\to Z{Z}^{*}$ are assumed to have the same signal-strength modification factor and are shown together as $VV$. All systematic uncertainties are correlated as in the nominal result.

Figure 17

Two-dimensional scans of the signal-strength modifiers for the processes (left) $t\overline{t}H,H\to b\overline{b}$ versus $t\overline{t}H,H\to W{W}^{*}/Z{Z}^{*}$ and (right) $t\overline{t}H,H\to \tau \tau$ versus $t\overline{t}H,H\to W{W}^{*}/Z{Z}^{*}$. The two signal strengths not appearing in each plot are profiled. The decays $H\to W{W}^{*}$ and $H\to Z{Z}^{*}$ are assumed to have the same signal-strength modification factor ${\mu }^{VV}$.

Figure 18

Allowed regions at 68% and 95% C.L. in the ${\kappa }_V\text{−}{\kappa }_F$ plane from the combination of all $t\overline{t}H$ channels. The Higgs boson is assumed to not couple to any particles beyond the Standard Model, and the $H\to \gamma \gamma$ and $H\to gg$ couplings are expressed in terms of ${\kappa }_F$ and ${\kappa }_V$.