Evidence for Higgs Boson Decays to the tau+tau- Final State with the ATLAS Detector

ATLAS-CONF-2013-108

28 November 2013

These preliminary results are superseded by the following paper:

HIGG-2013-32
ATLAS recommends to use the results from the paper.

Content

Preview

Main document (CDS record) - internal

pdf from CDS

Figures

Abstract

A search for the Higgs boson with a mass of about $125$ GeV decaying into a pair of $\tau$ leptons is performed with a data sample of proton-proton collisions, corresponding to an integrated luminosity of $\mathcal{L}=20.3$ fb$^{-1}$, collected with the ATLAS detector at the LHC at a centre-of-mass energy of $\sqrt{s}=8$ TeV. Final states in all $\tau$ decay combinations (both hadronic and leptonic) are examined. The observed (expected) deviation from the background-only hypothesis corresponds to a significance of 4.1 (3.2) standard deviations, and the measured signal strength is $\mu = 1.4 ^{+0.5}_{-0.4}$. This is evidence for the existence of $H\rightarrow\tau^+\tau^-$ decays, consistent with the Standard Model expectation for a Higgs boson with $m_H=125$ GeV.

Figures

Figure 01a:
Kinematic distributions for the τ_lepτ_lep channel after preselection: (a) E_T^miss, (b) m^vis_ττ, (c) p_T^H, and (d) Δη(j₁,j₂). The m^vis_ττ distribution shows a step at 75 GeV due to the difference in cuts between same flavour and different flavour event selection. The Δη(j₁,j₂) distribution is shown for events with at least two jets. Signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (91kB) eps (27kB) pdf (8kB)

Figure 01b:
Kinematic distributions for the τ_lepτ_lep channel after preselection: (a) E_T^miss, (b) m^vis_ττ, (c) p_T^H, and (d) Δη(j₁,j₂). The m^vis_ττ distribution shows a step at 75 GeV due to the difference in cuts between same flavour and different flavour event selection. The Δη(j₁,j₂) distribution is shown for events with at least two jets. Signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (91kB) eps (23kB) pdf (8kB)

Figure 01c:
Kinematic distributions for the τ_lepτ_lep channel after preselection: (a) E_T^miss, (b) m^vis_ττ, (c) p_T^H, and (d) Δη(j₁,j₂). The m^vis_ττ distribution shows a step at 75 GeV due to the difference in cuts between same flavour and different flavour event selection. The Δη(j₁,j₂) distribution is shown for events with at least two jets. Signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (84kB) eps (28kB) pdf (8kB)

Figure 01d:
Kinematic distributions for the τ_lepτ_lep channel after preselection: (a) E_T^miss, (b) m^vis_ττ, (c) p_T^H, and (d) Δη(j₁,j₂). The m^vis_ττ distribution shows a step at 75 GeV due to the difference in cuts between same flavour and different flavour event selection. The Δη(j₁,j₂) distribution is shown for events with at least two jets. Signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (86kB) eps (25kB) pdf (8kB)

Figure 02a:
Kinematic distributions for the τ_lepτ_had channel after preselection: (a) E_T^miss, (b) m^vis_ττ, (c) p_T^H, and (d) Δη(j₁,j₂). The Δη(j₁,j₂) distribution is shown for events with at least two jets. Signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (79kB) eps (22kB) pdf (7kB)

Figure 02b:
Kinematic distributions for the τ_lepτ_had channel after preselection: (a) E_T^miss, (b) m^vis_ττ, (c) p_T^H, and (d) Δη(j₁,j₂). The Δη(j₁,j₂) distribution is shown for events with at least two jets. Signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (83kB) eps (24kB) pdf (8kB)

Figure 02c:
Kinematic distributions for the τ_lepτ_had channel after preselection: (a) E_T^miss, (b) m^vis_ττ, (c) p_T^H, and (d) Δη(j₁,j₂). The Δη(j₁,j₂) distribution is shown for events with at least two jets. Signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (83kB) eps (23kB) pdf (8kB)

Figure 02d:
Kinematic distributions for the τ_lepτ_had channel after preselection: (a) E_T^miss, (b) m^vis_ττ, (c) p_T^H, and (d) Δη(j₁,j₂). The Δη(j₁,j₂) distribution is shown for events with at least two jets. Signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (75kB) eps (20kB) pdf (7kB)

Figure 03a:
Kinematic distributions for the τ_hadτ_had channel after preselection: (a) E_T^miss, (b) m^vis_ττ, (c) p_T^H, and (d) Δη(j₁,j₂). The Δη(j₁,j₂) distribution is shown for events with at least two jets. Signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (70kB) eps (20kB) pdf (7kB)

Figure 03b:
Kinematic distributions for the τ_hadτ_had channel after preselection: (a) E_T^miss, (b) m^vis_ττ, (c) p_T^H, and (d) Δη(j₁,j₂). The Δη(j₁,j₂) distribution is shown for events with at least two jets. Signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (80kB) eps (23kB) pdf (8kB)

Figure 03c:
Kinematic distributions for the τ_hadτ_had channel after preselection: (a) E_T^miss, (b) m^vis_ττ, (c) p_T^H, and (d) Δη(j₁,j₂). The Δη(j₁,j₂) distribution is shown for events with at least two jets. Signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (75kB) eps (23kB) pdf (8kB)

Figure 03d:
Kinematic distributions for the τ_hadτ_had channel after preselection: (a) E_T^miss, (b) m^vis_ττ, (c) p_T^H, and (d) Δη(j₁,j₂). The Δη(j₁,j₂) distribution is shown for events with at least two jets. Signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (77kB) eps (21kB) pdf (7kB)

Figure 04a:
Fake-factors used to derive estimates for multijet and W+jets backgrounds in the τ_lepτ_had channel. Factors are plotted as a function of the p_T of the τ_had candidate for the VBF and boosted categories: (a) for 1-prong τ_had candidates (b) for 3-prong τ_had candidates.

png (75kB) eps (15kB) pdf (7kB)

Figure 04b:
Fake-factors used to derive estimates for multijet and W+jets backgrounds in the τ_lepτ_had channel. Factors are plotted as a function of the p_T of the τ_had candidate for the VBF and boosted categories: (a) for 1-prong τ_had candidates (b) for 3-prong τ_had candidates.

png (78kB) eps (16kB) pdf (7kB)

Figure 05:
The Δη(τ_had,τ_had) distribution for the τ_hadτ_had channel after preselection. A fit to this distribution determines the starting values of the Z→τ⁺τ^- and multijet normalizations used in the global fit. The signal shape is shown multiplied by a factor of 50.

png (73kB) eps (21kB) pdf (7kB)

Figure 06a:
The b-tag jet multiplicity distributions, for jets with p_T>25 GeV, for the top-quark background control regions in the τ_lepτ_lep channel for the (a) VBF and (b) boosted categories. These figures use background predictions made without the global fit. The signal shapes are shown multiplied by a factor of 50.

png (67kB) eps (17kB) pdf (6kB)

Figure 06b:
The b-tag jet multiplicity distributions, for jets with p_T>25 GeV, for the top-quark background control regions in the τ_lepτ_lep channel for the (a) VBF and (b) boosted categories. These figures use background predictions made without the global fit. The signal shapes are shown multiplied by a factor of 50.

png (71kB) eps (19kB) pdf (6kB)

Figure 07a:
BDT score distributions for the Z→ℓℓ-enriched control region in the τ_lepτ_lep channel (top), Z→ττ-enriched control region in the τ_lepτ_had channel (middle), and mass sideband control region in the τ_hadτ_had channel (bottom), for the VBF (left) and Boosted (right) categories. The signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (83kB) eps (25kB) pdf (8kB)

Figure 07b:
BDT score distributions for the Z→ℓℓ-enriched control region in the τ_lepτ_lep channel (top), Z→ττ-enriched control region in the τ_lepτ_had channel (middle), and mass sideband control region in the τ_hadτ_had channel (bottom), for the VBF (left) and Boosted (right) categories. The signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (84kB) eps (26kB) pdf (8kB)

Figure 07c:
BDT score distributions for the Z→ℓℓ-enriched control region in the τ_lepτ_lep channel (top), Z→ττ-enriched control region in the τ_lepτ_had channel (middle), and mass sideband control region in the τ_hadτ_had channel (bottom), for the VBF (left) and Boosted (right) categories. The signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (73kB) eps (21kB) pdf (7kB)

Figure 07d:
BDT score distributions for the Z→ℓℓ-enriched control region in the τ_lepτ_lep channel (top), Z→ττ-enriched control region in the τ_lepτ_had channel (middle), and mass sideband control region in the τ_hadτ_had channel (bottom), for the VBF (left) and Boosted (right) categories. The signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (75kB) eps (22kB) pdf (7kB)

Figure 07e:
BDT score distributions for the Z→ℓℓ-enriched control region in the τ_lepτ_lep channel (top), Z→ττ-enriched control region in the τ_lepτ_had channel (middle), and mass sideband control region in the τ_hadτ_had channel (bottom), for the VBF (left) and Boosted (right) categories. The signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (70kB) eps (22kB) pdf (7kB)

Figure 07f:
BDT score distributions for the Z→ℓℓ-enriched control region in the τ_lepτ_lep channel (top), Z→ττ-enriched control region in the τ_lepτ_had channel (middle), and mass sideband control region in the τ_hadτ_had channel (bottom), for the VBF (left) and Boosted (right) categories. The signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (74kB) eps (23kB) pdf (7kB)

Figure 08a:
BDT distributions for the VBF (left) and boosted (right) category signal regions for the τ_lepτ_lep channel (top), τ_lepτ_had channel (middle), and τ_hadτ_had channel (bottom) channels. The Higgs boson signal (m_H = 125 GeV) is shown stacked with a signal strength of μ=1 (dashed line) and μ=1.4 (solid line). The background predictions come from the global fit (that gives μ=1.4). The size of the statistical and normalization systematics is indicated by the hashed band.

png (83kB) eps (23kB) pdf (8kB)

Figure 08b:
BDT distributions for the VBF (left) and boosted (right) category signal regions for the τ_lepτ_lep channel (top), τ_lepτ_had channel (middle), and τ_hadτ_had channel (bottom) channels. The Higgs boson signal (m_H = 125 GeV) is shown stacked with a signal strength of μ=1 (dashed line) and μ=1.4 (solid line). The background predictions come from the global fit (that gives μ=1.4). The size of the statistical and normalization systematics is indicated by the hashed band.

png (78kB) eps (23kB) pdf (8kB)

Figure 08c:
BDT distributions for the VBF (left) and boosted (right) category signal regions for the τ_lepτ_lep channel (top), τ_lepτ_had channel (middle), and τ_hadτ_had channel (bottom) channels. The Higgs boson signal (m_H = 125 GeV) is shown stacked with a signal strength of μ=1 (dashed line) and μ=1.4 (solid line). The background predictions come from the global fit (that gives μ=1.4). The size of the statistical and normalization systematics is indicated by the hashed band.

png (72kB) eps (21kB) pdf (7kB)

Figure 08d:
BDT distributions for the VBF (left) and boosted (right) category signal regions for the τ_lepτ_lep channel (top), τ_lepτ_had channel (middle), and τ_hadτ_had channel (bottom) channels. The Higgs boson signal (m_H = 125 GeV) is shown stacked with a signal strength of μ=1 (dashed line) and μ=1.4 (solid line). The background predictions come from the global fit (that gives μ=1.4). The size of the statistical and normalization systematics is indicated by the hashed band.

png (75kB) eps (23kB) pdf (7kB)

Figure 08e:
BDT distributions for the VBF (left) and boosted (right) category signal regions for the τ_lepτ_lep channel (top), τ_lepτ_had channel (middle), and τ_hadτ_had channel (bottom) channels. The Higgs boson signal (m_H = 125 GeV) is shown stacked with a signal strength of μ=1 (dashed line) and μ=1.4 (solid line). The background predictions come from the global fit (that gives μ=1.4). The size of the statistical and normalization systematics is indicated by the hashed band.

png (85kB) eps (24kB) pdf (8kB)

Figure 08f:
BDT distributions for the VBF (left) and boosted (right) category signal regions for the τ_lepτ_lep channel (top), τ_lepτ_had channel (middle), and τ_hadτ_had channel (bottom) channels. The Higgs boson signal (m_H = 125 GeV) is shown stacked with a signal strength of μ=1 (dashed line) and μ=1.4 (solid line). The background predictions come from the global fit (that gives μ=1.4). The size of the statistical and normalization systematics is indicated by the hashed band.

png (79kB) eps (23kB) pdf (8kB)

Figure 09:
The best-fit value for the signal strength μ in the individual channels and the combination. The total ±1σ uncertainty is indicated by the shaded green band, with the individual contributions from the statistical uncertainty (top, black), the total (experimental and theoretical) systematic uncertainty (middle, blue), and the theory uncertainty (bottom, red) on the signal cross section (from QCD scale, PDF, and branching ratios) shown by the error bars and printed in the central column.

png (129kB) eps (44kB) pdf (11kB)

Figure 10:
The likelihood dependence on the the signal strength parameter μ, where the y-axis is the deviation from the maximum likelihood: -2Δ ln(ℒ).

png (88kB) eps (12kB) pdf (5kB)

Figure 11:
Event yields as a function of log(S/B), where S (signal yield) and B (background yield) are taken from each event's bin in the BDT. All events in the BDTs are included. The predicted background is obtained from the global fit (with μ=1.4), and signal yields are shown for m_H=125 GeV, at μ=1 and μ=1.4 (the best-fit value). The background only distribution (dashed line), is obtained from the global fit, but fixing μ=0.

png (147kB) eps (26kB) pdf (7kB)

Figure 12a:
Distributions for m_ττ^MMC where events are weighted by ln(1+S/B) for all channels. These weights are determined by the signal and background predictions for each BDT bin. The bottom panel in each plot shows the difference between weighted data events and weighted background events (black points), compared to the weighted signal yields. The background predictions are obtained from the global fit with the m_H = 125 GeV signal hypothesis (μ=1.4). The m_H = 125 GeV signal is plotted with a solid red line, and, for comparison, the m_H = 110 GeV (blue) and m_H = 150 GeV (green) signals are also shown. (a) Shows all signal strengths set to the Standard Model expectation while (b) shows the signal strengths set to their best fit values.

png (111kB) eps (30kB) pdf (11kB)

Figure 12b:
Distributions for m_ττ^MMC where events are weighted by ln(1+S/B) for all channels. These weights are determined by the signal and background predictions for each BDT bin. The bottom panel in each plot shows the difference between weighted data events and weighted background events (black points), compared to the weighted signal yields. The background predictions are obtained from the global fit with the m_H = 125 GeV signal hypothesis (μ=1.4). The m_H = 125 GeV signal is plotted with a solid red line, and, for comparison, the m_H = 110 GeV (blue) and m_H = 150 GeV (green) signals are also shown. (a) Shows all signal strengths set to the Standard Model expectation while (b) shows the signal strengths set to their best fit values.

png (113kB) eps (30kB) pdf (11kB)

Figure 13:
Likelihood contours for the H→ττ channel in the (μ_ggF× B/B_SM, μ_VBF+VH× B/B_SM) plane are shown for the 68% and 95% CL by dashed and solid lines, respectively, for m_H=125 GeV. The SM expectation and the one corresponding to background-only hypothesis are shown by a filled plus and an open plus symbol, respectively. The best-fit to the data is shown for the case when both the μ_ggF and μ_VBF+VH are unconstrained.

png (79kB) eps (23kB) pdf (9kB)

Figure 14a:
BDT input variables for the τ_lepτ_had VBF category obtained in the signal region. The following distributions are shown: (a) m_T, (b) p_T^Total, (c) Δη(j₁,j₂) and (d) ΔR(τ,τ). The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit. The lowest bin in the p_T^Total distribution acts as an underflow bin, and includes events with p_T^Total < 30 GeV.

png (77kB) eps (19kB) pdf (7kB)

Figure 14b:
BDT input variables for the τ_lepτ_had VBF category obtained in the signal region. The following distributions are shown: (a) m_T, (b) p_T^Total, (c) Δη(j₁,j₂) and (d) ΔR(τ,τ). The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit. The lowest bin in the p_T^Total distribution acts as an underflow bin, and includes events with p_T^Total < 30 GeV.

png (73kB) eps (20kB) pdf (7kB)

Figure 14c:
BDT input variables for the τ_lepτ_had VBF category obtained in the signal region. The following distributions are shown: (a) m_T, (b) p_T^Total, (c) Δη(j₁,j₂) and (d) ΔR(τ,τ). The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit. The lowest bin in the p_T^Total distribution acts as an underflow bin, and includes events with p_T^Total < 30 GeV.

png (91kB) eps (25kB) pdf (8kB)

Figure 14d:
BDT input variables for the τ_lepτ_had VBF category obtained in the signal region. The following distributions are shown: (a) m_T, (b) p_T^Total, (c) Δη(j₁,j₂) and (d) ΔR(τ,τ). The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit. The lowest bin in the p_T^Total distribution acts as an underflow bin, and includes events with p_T^Total < 30 GeV.

png (98kB) eps (26kB) pdf (9kB)

Figure 15a:
BDT input variables for the τ_lepτ_had VBF category obtained in the signal region. The following distributions are shown: (a) η_j₁× η_j₂, (b) ℓ η centrality, (c) E_T^missφ centrality and (d) m_j₁,j₂. The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit.

png (96kB) eps (23kB) pdf (8kB)

Figure 15b:
BDT input variables for the τ_lepτ_had VBF category obtained in the signal region. The following distributions are shown: (a) η_j₁× η_j₂, (b) ℓ η centrality, (c) E_T^missφ centrality and (d) m_j₁,j₂. The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit.

png (92kB) eps (24kB) pdf (9kB)

Figure 15c:
BDT input variables for the τ_lepτ_had VBF category obtained in the signal region. The following distributions are shown: (a) η_j₁× η_j₂, (b) ℓ η centrality, (c) E_T^missφ centrality and (d) m_j₁,j₂. The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit.

png (85kB) eps (21kB) pdf (8kB)

Figure 15d:
BDT input variables for the τ_lepτ_had VBF category obtained in the signal region. The following distributions are shown: (a) η_j₁× η_j₂, (b) ℓ η centrality, (c) E_T^missφ centrality and (d) m_j₁,j₂. The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit.

png (90kB) eps (22kB) pdf (8kB)

Figure 16a:
BDT input variables for the τ_hadτ_had VBF category obtained in the signal region. The following distributions are shown: (a) τ₁ η centrality, (b) τ₂ η centrality, (c) Δη(j₁,j₂) and (d) Δ R(τ,τ). The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit.

png (83kB) eps (23kB) pdf (8kB)

Figure 16b:
BDT input variables for the τ_hadτ_had VBF category obtained in the signal region. The following distributions are shown: (a) τ₁ η centrality, (b) τ₂ η centrality, (c) Δη(j₁,j₂) and (d) Δ R(τ,τ). The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit.

png (84kB) eps (23kB) pdf (8kB)

Figure 16c:
BDT input variables for the τ_hadτ_had VBF category obtained in the signal region. The following distributions are shown: (a) τ₁ η centrality, (b) τ₂ η centrality, (c) Δη(j₁,j₂) and (d) Δ R(τ,τ). The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit.

png (76kB) eps (21kB) pdf (7kB)

Figure 16d:
BDT input variables for the τ_hadτ_had VBF category obtained in the signal region. The following distributions are shown: (a) τ₁ η centrality, (b) τ₂ η centrality, (c) Δη(j₁,j₂) and (d) Δ R(τ,τ). The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit.

png (75kB) eps (21kB) pdf (7kB)

Figure 17a:
BDT input variables for the τ_hadτ_had VBF category obtained in the signal region. The following distributions are shown: (a) η_j₁× η_j₂, (b) p_T^Total, (c) E_T^missφ centrality and (d) m_j₁,j₂. The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit.

png (74kB) eps (20kB) pdf (7kB)

Figure 17b:
BDT input variables for the τ_hadτ_had VBF category obtained in the signal region. The following distributions are shown: (a) η_j₁× η_j₂, (b) p_T^Total, (c) E_T^missφ centrality and (d) m_j₁,j₂. The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit.

png (86kB) eps (24kB) pdf (8kB)

Figure 17c:
BDT input variables for the τ_hadτ_had VBF category obtained in the signal region. The following distributions are shown: (a) η_j₁× η_j₂, (b) p_T^Total, (c) E_T^missφ centrality and (d) m_j₁,j₂. The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit.

png (78kB) eps (23kB) pdf (8kB)

Figure 17d:
BDT input variables for the τ_hadτ_had VBF category obtained in the signal region. The following distributions are shown: (a) η_j₁× η_j₂, (b) p_T^Total, (c) E_T^missφ centrality and (d) m_j₁,j₂. The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit.

png (86kB) eps (23kB) pdf (8kB)

Figure 18a:
BDT input variables for the τ_lepτ_lep VBF category obtained in the signal region. The following distributions are shown: (a) Δη(j₁,j₂), (b) Δ R(τ,τ), (c) min(Δ η_ℓ₁ℓ_₂,jets), (d) ℓ η centrality, (e) j₃ η centrality and (f) m_j₁,j₂. The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit.

png (83kB) eps (22kB) pdf (7kB)

Figure 18b:
BDT input variables for the τ_lepτ_lep VBF category obtained in the signal region. The following distributions are shown: (a) Δη(j₁,j₂), (b) Δ R(τ,τ), (c) min(Δ η_ℓ₁ℓ_₂,jets), (d) ℓ η centrality, (e) j₃ η centrality and (f) m_j₁,j₂. The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit.

png (79kB) eps (21kB) pdf (7kB)

Figure 18c:
BDT input variables for the τ_lepτ_lep VBF category obtained in the signal region. The following distributions are shown: (a) Δη(j₁,j₂), (b) Δ R(τ,τ), (c) min(Δ η_ℓ₁ℓ_₂,jets), (d) ℓ η centrality, (e) j₃ η centrality and (f) m_j₁,j₂. The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit.

png (85kB) eps (23kB) pdf (7kB)

Figure 18d:
BDT input variables for the τ_lepτ_lep VBF category obtained in the signal region. The following distributions are shown: (a) Δη(j₁,j₂), (b) Δ R(τ,τ), (c) min(Δ η_ℓ₁ℓ_₂,jets), (d) ℓ η centrality, (e) j₃ η centrality and (f) m_j₁,j₂. The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit.

png (83kB) eps (23kB) pdf (7kB)

Figure 18e:
BDT input variables for the τ_lepτ_lep VBF category obtained in the signal region. The following distributions are shown: (a) Δη(j₁,j₂), (b) Δ R(τ,τ), (c) min(Δ η_ℓ₁ℓ_₂,jets), (d) ℓ η centrality, (e) j₃ η centrality and (f) m_j₁,j₂. The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit.

png (85kB) eps (23kB) pdf (7kB)

Figure 18f:
BDT input variables for the τ_lepτ_lep VBF category obtained in the signal region. The following distributions are shown: (a) Δη(j₁,j₂), (b) Δ R(τ,τ), (c) min(Δ η_ℓ₁ℓ_₂,jets), (d) ℓ η centrality, (e) j₃ η centrality and (f) m_j₁,j₂. The signal shapes are shown multiplied by a factor of 20. These figures use background predictions made without the global fit.

png (86kB) eps (25kB) pdf (7kB)

Figure 19a:
Distributions for m_ττ^MMC normalized to unit area for Z→τ⁺τ^- , from τ-embedded Z→μ⁺μ^- data, and H→τ⁺τ^- events in the (a) τ_lepτ_lep, (b) τ_lepτ_had and (c) τ_hadτ_had channels for events that pass preselection requirements and have at least 1 jet. Full width at half maximum (FWHM) values for the τ_lepτ_lep distributions are: 42 GeV for H→τ⁺τ^- and 39 GeV for Z→τ⁺τ^-. FWHM values for the τ_lepτ_had distributions are: 47 GeV for H→τ⁺τ^- and 33 GeV for Z→τ⁺τ^- . FWHM values for the τ_hadτ_had distributions are: 38 GeV for H→τ⁺τ^- and 26 GeV for Z→τ⁺τ^-.

png (53kB) eps (11kB) pdf (5kB)

Figure 19b:
Distributions for m_ττ^MMC normalized to unit area for Z→τ⁺τ^- , from τ-embedded Z→μ⁺μ^- data, and H→τ⁺τ^- events in the (a) τ_lepτ_lep, (b) τ_lepτ_had and (c) τ_hadτ_had channels for events that pass preselection requirements and have at least 1 jet. Full width at half maximum (FWHM) values for the τ_lepτ_lep distributions are: 42 GeV for H→τ⁺τ^- and 39 GeV for Z→τ⁺τ^-. FWHM values for the τ_lepτ_had distributions are: 47 GeV for H→τ⁺τ^- and 33 GeV for Z→τ⁺τ^- . FWHM values for the τ_hadτ_had distributions are: 38 GeV for H→τ⁺τ^- and 26 GeV for Z→τ⁺τ^-.

png (54kB) eps (11kB) pdf (5kB)

Figure 19c:
Distributions for m_ττ^MMC normalized to unit area for Z→τ⁺τ^- , from τ-embedded Z→μ⁺μ^- data, and H→τ⁺τ^- events in the (a) τ_lepτ_lep, (b) τ_lepτ_had and (c) τ_hadτ_had channels for events that pass preselection requirements and have at least 1 jet. Full width at half maximum (FWHM) values for the τ_lepτ_lep distributions are: 42 GeV for H→τ⁺τ^- and 39 GeV for Z→τ⁺τ^-. FWHM values for the τ_lepτ_had distributions are: 47 GeV for H→τ⁺τ^- and 33 GeV for Z→τ⁺τ^- . FWHM values for the τ_hadτ_had distributions are: 38 GeV for H→τ⁺τ^- and 26 GeV for Z→τ⁺τ^-.

png (54kB) eps (10kB) pdf (5kB)

Figure 20a:
Distributions of m_ττ^MMC for the following categories: (a) τ_lepτ_lep VBF, (b) τ_lepτ_lep boosted, (c) τ_lepτ_had VBF, (d) τ_lepτ_had boosted, (e) τ_hadτ_had VBF and (f) τ_hadτ_had boosted. The background and signal predictions are shown from the global fit.

png (97kB) eps (25kB) pdf (9kB)

Figure 20b:
Distributions of m_ττ^MMC for the following categories: (a) τ_lepτ_lep VBF, (b) τ_lepτ_lep boosted, (c) τ_lepτ_had VBF, (d) τ_lepτ_had boosted, (e) τ_hadτ_had VBF and (f) τ_hadτ_had boosted. The background and signal predictions are shown from the global fit.

png (89kB) eps (24kB) pdf (8kB)

Figure 20c:
Distributions of m_ττ^MMC for the following categories: (a) τ_lepτ_lep VBF, (b) τ_lepτ_lep boosted, (c) τ_lepτ_had VBF, (d) τ_lepτ_had boosted, (e) τ_hadτ_had VBF and (f) τ_hadτ_had boosted. The background and signal predictions are shown from the global fit.

png (90kB) eps (29kB) pdf (9kB)

Figure 20d:
Distributions of m_ττ^MMC for the following categories: (a) τ_lepτ_lep VBF, (b) τ_lepτ_lep boosted, (c) τ_lepτ_had VBF, (d) τ_lepτ_had boosted, (e) τ_hadτ_had VBF and (f) τ_hadτ_had boosted. The background and signal predictions are shown from the global fit.

png (80kB) eps (25kB) pdf (8kB)

Figure 20e:
Distributions of m_ττ^MMC for the following categories: (a) τ_lepτ_lep VBF, (b) τ_lepτ_lep boosted, (c) τ_lepτ_had VBF, (d) τ_lepτ_had boosted, (e) τ_hadτ_had VBF and (f) τ_hadτ_had boosted. The background and signal predictions are shown from the global fit.

png (91kB) eps (25kB) pdf (8kB)

Figure 20f:
Distributions of m_ττ^MMC for the following categories: (a) τ_lepτ_lep VBF, (b) τ_lepτ_lep boosted, (c) τ_lepτ_had VBF, (d) τ_lepτ_had boosted, (e) τ_hadτ_had VBF and (f) τ_hadτ_had boosted. The background and signal predictions are shown from the global fit.

png (84kB) eps (23kB) pdf (8kB)

Figure 21a:
Distributions for m_ττ^MMC where events are weighted by ln(1+S/B) for the τ_lepτ_lep (a), τ_lepτ_had (b), and τ_hadτ_had (c) channels. These weights are determined by the signal and background predictions for each BDT bin. Signal is shown stacked on background, with both predictions coming from the global fit yielding a signal strength of μ = 1.4. The bottom panel shows the difference between weighted data events and weighted background events (black points), compared to the weighted signal yields. The m_H = 125 GeV signal is plotted with a solid red line, and, for comparison, the m_H = 110 GeV (blue) and m_H = 150 GeV (green) signals are plotted with a signal strength set to the observed values from the global fit, for those respective signal hypotheses.

png (109kB) eps (29kB) pdf (10kB)

Figure 21b:
Distributions for m_ττ^MMC where events are weighted by ln(1+S/B) for the τ_lepτ_lep (a), τ_lepτ_had (b), and τ_hadτ_had (c) channels. These weights are determined by the signal and background predictions for each BDT bin. Signal is shown stacked on background, with both predictions coming from the global fit yielding a signal strength of μ = 1.4. The bottom panel shows the difference between weighted data events and weighted background events (black points), compared to the weighted signal yields. The m_H = 125 GeV signal is plotted with a solid red line, and, for comparison, the m_H = 110 GeV (blue) and m_H = 150 GeV (green) signals are plotted with a signal strength set to the observed values from the global fit, for those respective signal hypotheses.

png (115kB) eps (31kB) pdf (11kB)

Figure 21c:
Distributions for m_ττ^MMC where events are weighted by ln(1+S/B) for the τ_lepτ_lep (a), τ_lepτ_had (b), and τ_hadτ_had (c) channels. These weights are determined by the signal and background predictions for each BDT bin. Signal is shown stacked on background, with both predictions coming from the global fit yielding a signal strength of μ = 1.4. The bottom panel shows the difference between weighted data events and weighted background events (black points), compared to the weighted signal yields. The m_H = 125 GeV signal is plotted with a solid red line, and, for comparison, the m_H = 110 GeV (blue) and m_H = 150 GeV (green) signals are plotted with a signal strength set to the observed values from the global fit, for those respective signal hypotheses.

png (108kB) eps (29kB) pdf (10kB)

Figure 22:
Distribution of the q_μ=0 for the background only (blue) and signal+background hypothesis (red), in the asymptotic approximation. The observed value (dotted line) obtained on data is also shown. The probability of the background only hypothesis (p_b) and the probability of the signal plus background hypothesis (p_s+b), are computed as the integral of f(q_μ=0 | hyp) over q_μ=0>q^obs_μ=0.

png (85kB) eps (37kB) pdf (15kB)

Figure 23a:
BDT score distributions for the Z→ττ-enriched (a, c) and top-quark enriched (b, d) control regions for the VBF (a, b) and Boosted (c, d) categories in the τ_lepτ_lep channel. The signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (79kB) eps (24kB) pdf (8kB)

Figure 23b:
BDT score distributions for the Z→ττ-enriched (a, c) and top-quark enriched (b, d) control regions for the VBF (a, b) and Boosted (c, d) categories in the τ_lepτ_lep channel. The signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (101kB) eps (26kB) pdf (9kB)

Figure 23c:
BDT score distributions for the Z→ττ-enriched (a, c) and top-quark enriched (b, d) control regions for the VBF (a, b) and Boosted (c, d) categories in the τ_lepτ_lep channel. The signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (84kB) eps (24kB) pdf (8kB)

Figure 23d:
BDT score distributions for the Z→ττ-enriched (a, c) and top-quark enriched (b, d) control regions for the VBF (a, b) and Boosted (c, d) categories in the τ_lepτ_lep channel. The signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (95kB) eps (26kB) pdf (9kB)

Figure 24a:
BDT score distributions for the top-quark enriched (a, c) and W-enriched control regions (b, d) for the VBF (a, b) and Boosted (c, d) categories in the τ_lepτ_had channel. The signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (73kB) eps (19kB) pdf (7kB)

Figure 24b:
BDT score distributions for the top-quark enriched (a, c) and W-enriched control regions (b, d) for the VBF (a, b) and Boosted (c, d) categories in the τ_lepτ_had channel. The signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (75kB) eps (19kB) pdf (7kB)

Figure 24c:
BDT score distributions for the top-quark enriched (a, c) and W-enriched control regions (b, d) for the VBF (a, b) and Boosted (c, d) categories in the τ_lepτ_had channel. The signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (74kB) eps (20kB) pdf (7kB)

Figure 24d:
BDT score distributions for the top-quark enriched (a, c) and W-enriched control regions (b, d) for the VBF (a, b) and Boosted (c, d) categories in the τ_lepτ_had channel. The signal shapes are shown multiplied by a factor of 50. These figures use background predictions made without the global fit.

png (76kB) eps (20kB) pdf (7kB)

Figure 25a:
Event yields as a function of log(S/B) in all signal regions bins in the τ_lepτ_lep (a), τ_lepτ_had (b), τ_hadτ_had (c) channels. The predicted background, obtained from the global fit (μ=1.4), and signal yields, for m_H=125 GeV, at μ=1 and μ=1.4 are shown. The background only distribution (dashed line), is obtained from the global fit, but fixing μ=0.

png (146kB) eps (24kB) pdf (7kB)

Figure 25b:
Event yields as a function of log(S/B) in all signal regions bins in the τ_lepτ_lep (a), τ_lepτ_had (b), τ_hadτ_had (c) channels. The predicted background, obtained from the global fit (μ=1.4), and signal yields, for m_H=125 GeV, at μ=1 and μ=1.4 are shown. The background only distribution (dashed line), is obtained from the global fit, but fixing μ=0.

png (146kB) eps (25kB) pdf (7kB)

Figure 25c:
Event yields as a function of log(S/B) in all signal regions bins in the τ_lepτ_lep (a), τ_lepτ_had (b), τ_hadτ_had (c) channels. The predicted background, obtained from the global fit (μ=1.4), and signal yields, for m_H=125 GeV, at μ=1 and μ=1.4 are shown. The background only distribution (dashed line), is obtained from the global fit, but fixing μ=0.

png (141kB) eps (24kB) pdf (7kB)

Figure 26:
The measured production strengths normalized to the SM expectations, for the H→γγ, H→ZZ(*)→4ℓ, H→WW(*)→ℓνℓν final states and their combination together with the preliminary signal strength measurements for the H→bbar and H→ττ final states. The measured production strengths are at m_H=125.5 GeV, except for Hτoτauτau which is at m_H=125 GeV. The best-fit values are shown by the solid vertical lines. The total ±1σ uncertainty is indicated by the shaded band, with the individual contributions from the statistical uncertainty (top), the total (experimental and theoretical) systematic uncertainty (middle), and the theory uncertainty (bottom) on the signal cross section (from QCD scale, PDF, and branching ratios) shown as superimposed error bars.

png (148kB) eps (33kB) pdf (9kB)

Figure 27:
Display of an event selected by the H→τ_lepτ_lep channel in the VBF category, where one τ decays to a muon and the other to an electron. The electron is indicated by a blue track and the muon indicated by a red track. The approximately horizontal dashed line in the R-φ view represents the direction of the E_T^miss vector, and there are two VBF jets marked with turquoise cones. The muon p_T is 53 GeV, the electron p_T is 34 GeV, E_T^miss=102 GeV, m_j₁,j₂=1.04 TeV and m_ττ^MMC=127 GeV and BDT=0.97. The S/B ratio in the BDT score bin of this event is 0.42.

png (1MB)

Figure 28:
Display of an event selected by the H→τ_lepτ_had channel in the VBF category, where one τ decays to an electron. The hadronically decaying τ lepton (1 prong decay) is indicated by a green track and the yellow cluster. The electron is indicated by the blue track match to the green cluster. The approximately horizontal dashed line in the R-φ view represents the direction of the E_T^miss vector, and there are two VBF jets marked with turquoise cones. The electron p_T is 56 GeV, the τ_had p_T is 27 GeV, E_T^miss=113 GeV, m_j₁,j₂=1.53 TeV, m_ττ^MMC=129 GeV, and the BDT score is 0.99. The S/B ratio in the BDT score bin of this event is 1.0.

png (992kB)

Figure 29:
Display of an event selected by the H→τ_hadτ_had channel in the VBF category, where the two τ decays into hadrons. The hadronically decaying τ leptons are indicated by green tracks. The dashed line in the lower left quadrant of the R-φ view represents the direction of the E_T^miss vector, and there are two VBF jets marked with turquoise cones. The leading τ_had p_T is 122 GeV, the sub-leading τ_had p_T is 67 GeV, E_T^miss=72 GeV, m_j₁,j₂=1.02 TeV and m_ττ^MMC=126 GeV and BDT=1.0. The S/B ratio in the BDT score bin of this event is 0.67.

png (895kB)

© 2013 CERN for the benefit of the ATLAS Collaboration. Reproduction of the article, figures and tables on this page is allowed as specified in the CC-BY-4.0 license.
2024-04-19 01:08:41