DOI 10.1393/ncc/i2017-17182-6 Colloquia: LaThuile 2017
SM Higgs boson measurements at CMS
M. Malberti on behalf of the CMS Collaboration
INFN and Universit`a Milano Bicocca - Milano, Italy
received 16 September 2017
Summary. — Measurements of the Higgs boson performed by the CMS experiment at the LHC are presented. A selection of preliminary results based on proton-proton collisions data collected at a center of mass energy√s = 13 TeV and corresponding
to an integrated luminosity of 12.9 fb−1is reported.
1. – Introduction
After the discovery of the Higgs boson by the ATLAS and CMS Collaborations [1, 2], a wide range of production and decay channels has been studied using the Run 1 data set at √s = 7 and 8 TeV to characterize the new observed particle. Measurements of
its couplings and properties are found to be consistent with the Standard Model (SM) expectations, the mass is measured with 0.2% precision and only a small excess in the t¯tH channel has been observed [3].
Preliminary results at √s = 13 TeV are presented here, based on the analysis of
proton-proton collisions data collected by the CMS experiment [4] in 2016 and corre-sponding to and integrated luminosity of 12.9 fb−1.
2. – H→ZZ→4l
Higgs boson decays to four leptons are selected from events with two pairs of opposite-sign, same-flavor well reconstructed and isolated leptons (electrons or muons). This chan-nel is characterized by a fully reconstructed mass peak with large signal-over-background (fig. 1 (left)). The analysis strategy is based on the definition of exclusive categories, with selections on kinematic discriminants defined using matrix element methods (MEM) and on the number of (b-)jets and additional leptons. The main backgrounds are from
qq→ZZ∗and gg→ZZ∗processes and from fake leptons from Z+jets, Z+b¯b, t¯t processes. The significance observed with the 12.9 fb−1data set is 6.2σ, where 6.5σ are expected. The best-fit signal strength μ = σ/σSM at the Run 1 measured mass mH= 125.09 GeV
is μ = 0.99+0.33−0.26. The signal strength measured per production mechanism, the fiducial cross section (fig. 2 (left)) and differential cross section measurements as a function of
c
CERN on behalf of the CMS Collaboration
(GeV) l 4 m 70 80 90 100 110 120 130 140 150 160 170 Events / 4 GeV 0 5 10 15 20 25 30 35 40 45 Data H(125) * γ ZZ, Z → q q * γ ZZ, Z → gg Z+X (13 TeV) -1 12.9 fb CMSPreliminary Events / GeV 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 Data S+B fit B component σ 1 ± σ 2 ± All categories =0.95 μ =126.0 GeV, H m --Preliminary CMS 12.9 fb-1 (13TeV) γ γ → H (GeV) γ γ m 100 110 120 130 140 150 160 170 180 100 − 0 100 200 B component subtracted
Fig. 1. – Four leptons invariant mass distribution in the H→ ZZ channel [5] (left) and di-photon invariant mass in the H→ γγ channel [6] (right).
(TeV) s 7 8 9 10 11 12 13 14 [fb] fid σ 0 1 2 3 4 5 sys. unc.) ⊕ Data (stat. Systematic uncertainty Model dependence H) → LO gg 3 = 125 GeV, N H Standard model (m (13 TeV) -1 (8 TeV), 12.9 fb -1 (7 TeV), 19.7 fb -1 5.1 fb CMSPreliminary 4l) + X → (H → pp (H) [fb/GeV] T /dp fid σ d 10-3 -2 10 -1 10 1 (13 TeV) -1 12.9 fb CMSPreliminary sys. unc.) ⊕
Data (stat. gg→H (POWHEG+JHUGen) + XH
Systematic uncertainty Model dependence XH = VBF + VH + ttH (H) [GeV] T p 0 20 40 60 80 100 120 140 160 180 200 Ratio to POWHEG 0.20 0.4 0.6 0.81 1.2 1.4 1.6 1.82 2.2 2.4
Fig. 2. – Measurements of the fiducial cross section as a function of the center of mass energy (left) and differential cross section as a function of the Higgs boson pT(right) in the H→ ZZ → 4l
decay channel [5].
the Higgs boson transverse momentum (fig. 2 (right)) and number of jets in the events are compatible with the SM expectations [5].
3. – H→ γγ
Despite its small branching ratio (∼ 0.23% for mH= 125 GeV), the H→ γγ decay channel is characterized by a clean experimental signature, with two high transverse momentum isolated photons, which allow high precision for mass reconstruction.
To achieve the maximum sensitivity, events are classified in exclusive categories ex-ploiting their different mass resolution and signal-over-background ratio. The event in-formation, including the kinematics, photon quality and mass resolution, is combined in a multivariate classifier, which is built in such a way to be mass independent and to have high values for events with good di-photon mass resolution and high probability of being signal rather than background. Additional categorization is performed to select events with jets and leptons targeting specific Higgs production modes (namely VBF and t¯tH associated production).
μ 2 − 0 2 4 6 8 -1.2 +1.5 1.91 ttH μ -0.8 +0.9 1.61 VBF μ -0.23 +0.25 0.77 ggH μ 0.18 − 0.21 + = 0.95 combined μ Profiled H m σ 1 ± Combined σ 1 ± Per category SM μ = μ Preliminary CMS γ γ → H TeV) (13 -1 12.9 fb = 1 VH μ (TeV) s 7 8 9 10 11 12 13 14 (fb) fid. σ 20 30 40 50 60 70 80 90 100 Preliminary CMS (13 TeV) -1 (8 TeV) + 12.9 fb -1 19.7 fb γ γ → H ) H best-fit m Data ( syst. uncertainty ) =125.09 GeV H m SM (
- norm. LHC Higgs XSWG YR4
MC@NLO
A - acc.
Fig. 3. – Signal strength per production mechanism (left) and fiducial cross section as a function of the center of mass energy (right) measured in the H→ γγ decay channel [6].
A clear signal is observed in the di-photon channel (fig. 1 (right)), with a significance of 6.1σ. The best-fit signal strength is μ = 0.95+0.21−0.18(fig. 3 (left)) and the best-fit value of the fiducial cross section is found to be σf id = 69+18
−22fb, where the SM theoretical
prediction is 73.8± 3.8 fb (fig. 3 (right)). All the measurements are consistent with the expectations from a SM Higgs boson [6].
4. – t¯tH searches
4.1. t¯tH multileptons. – The search for t¯tH in multileptons targets a signature with H decays to W+W−/ZZ/τ+τ− final states accompanied by additional products from t¯t decays. Events with multiple leptons and jets are selected and further categorization is performed based on the lepton flavour and charge, number of b-jets and hadronic τ decays. The main irreducible backgrounds are represented by t¯tV and di-bosons, while
Events 0 20 40 60 80 100 120 140 160 Post-fit, Dataμ ttH TTW TTZ Fakes Flips Preliminary CMS 12.9 fb-1 (13 TeV) BDT (ttH,tt/ttV) bin 1 2 3 4 5 6 7 Data/Pred. 0.5 1.0 1.5 2.0 SM σ / σ = μ Best fit 2 − −1 0 1 2 3 4 5 6 1.3 + 1.2 − = 2.4 μ h τ , no μ e 1.3 + 1.1 − = 2.6 μ h τ , no μ μ 2.4 + 2.3 − = 1.2 μ h τ ee, no 1.4 + 1.2 − = 0.0 μ h τ 2l, 1 1.4 + 1.2 − = 2.5 μ 3l (13 TeV) -1 2.3+12.9 fb Preliminary CMS = 125 GeV H m 0.8 + 0.7 − = 2.0 μ combined
Fig. 4. – Example of BDT classifier output in the bins used for signal extraction in t¯tH mul-tilepton searches, for the same-sign dilepton channel (left); combined and per category best-fit signal strength for the combined 2015+2016 analysis (right) [7].
Events 20 40 60 80 100 120 140 data 15 × ttH +LF t t tt+cc +b t t tt+2b b +b t t single-t +V t t V+jets diboson (13 TeV) -1 11.4 - 12.9 fb Preliminary CMS 3 jets, 3 b-tags dilepton pre-fit expectation BDT discriminant 1.0 − −0.8 −0.6 −0.4 −0.2 0.0 0.2 0.4 Data/Bkg. 0.5 1.0 1.5 = 125 GeV H at m SM σ / σ = μ 95% CL limit on 1 10 Combined Lepton+jets Dilepton Preliminary CMS (13 TeV) -1 11.4 - 12.9 fb σ 1 ± Expected σ 2 ± Expected =1) injected μ H( t t Observed
Fig. 5. – Example of BDT classifier output used for signal extraction, for the dilepton category with 3 jets and 3 b-jets (left) and upper limit of σ/σSM in the t¯tH(H→ b¯b) search channel
(right) [8].
reducible backgrounds by non-prompt leptons from t¯t events. Boosted decision trees (BDT) are trained to separate the t¯tH signal from the generic t¯t background and from the irreducible t¯tV background, exploiting topological and kinematic variables, such as jet mutiplicity, lepton/jet angular sepration, missing transverse energy, leptons pT. A signal is observed with 3.2σ significance (where 1.7σ is expected) from the combined analysis of the 2015 (2.3 fb−1) and 2016 (12.9 fb−1) data sets, with a best-fit signal strength μ = 2.0+0.8−0.7(fig. 4) [7].
4.2. t¯tH(H→b¯b). – The t¯tH search through H→ b¯b decays exploits the large branching fraction of the Higgs to b¯b pairs (58%), but on the other hand is affected by large backgrounds. Two channels are considered for this search: the lepton+jets channel, with one lepton and at least four jets, and the dilepton channel, with two opposite sign leptons
Events / GeV 0 1 2 3 4 5 6 7 8 9 Data S+B fit B component σ 1 ± σ 2 ± TTH Leptonic Tag =0.95 μ =126.0 GeV, H m --Preliminary CMS 12.9 fb-1 (13TeV) γ γ → H (GeV) γ γ m 100 110 120 130 140 150 160 170 180 2 − 0 2 4 6 B component subtracted Events / GeV 0 2 4 6 8 10 12 Data S+B fit B component σ 1 ± σ 2 ± TTH Hadronic Tag =0.95 μ =126.0 GeV, H m --Preliminary CMS 12.9 fb-1 (13TeV) γ γ → H (GeV) γ γ m 100 110 120 130 140 150 160 170 180 4 − 2 − 0 2 4 6 8 10 B component subtracted
Fig. 7. – Projections of the precision on the signal strengths and fiducial cross section in the H→ γγ channel [9].
and at least two jets. Further classification is based on the number of jets and b-jets in the event. Sub-categories are finally defined based on multivariate and MEM discriminants. The MEM discriminant is optimized to separate the signal from the irreducible t¯tb¯b background. A combined fit of multivariate discriminant distributions in all categories results in an observed (expected) upper limit of σ/σSM < 1.5 (1.7) at the 95% confidence
level (fig. 5) [8].
4.3. t¯tH(H→ γγ). – The search for t¯tH production with H→ γγ decays follows the analysis strategy of the H→ γγ analysis described in sect. 3. Events are then categorized depending on the decay of the t¯t pair in two categories: a leptonic category, with two photons, at least one lepton (electron or muon) and three jets and a hadronic category, with two photons, at least five jets and no leptons. In each category, at least one of the jets must be b-tagged. Figure 6 shows the di-photon invariant mass distribution in the two categories. The best-fit signal strength for t¯tH production in the di-photon decay channel is μ = 1.91+1.5−1.2[6].
Fig. 8. – Projections of the precision on the signal strengths and fiducial cross section in the
5. – Projections
The results obtained from the analysis of 13 TeV data are extrapolated to larger data sets of 300 fb−1and 3000 fb−1, considering an upgraded CMS detector for the HL-LHC. Extrapolations are studied under different scenarios for the systematic uncertainties as-sumed in the measurements, which are either kept constant with the intergrated lumi-nosity or scaled down (theoretical uncertainties by a factor 2 and experimental ones by the square root of the integrated luminosity until they reach a defined lower limit based on estimates of the achievable accuracy with the upgraded detector) [9]. A significant improvement with respect to the current measurements is expected in the precision on the couplings, fiducial and differential cross sections, as shown in fig. 7 and fig. 8.
6. – Conclusions and outlook
A selection of preliminary Higgs boson measurements performed at CMS using 12.9 fb−1 of 13 TeV pp collisions data has been presented. The sensitivity reached with this data set is already close to the Run 1 one and measurements are found to be largely compatible with SM expectations. All the measurements are being updated analysing the entire 2016 data set (about 36 fb−1) and substantial improvements are foreseen. LHC is expected to deliver more than 100 fb−1by the end of Run 2, allowing a further increase on the precision of the Higgs properties measurements.
REFERENCES
[1] ATLAS Collaboration, Phys. Lett. B, 716 (2012) 129. [2] CMS Collaboration, Phys. Lett. B, 716 (2012) 3061. [3] ATLAS and Collaborations CMS, JHEP, 08 (2016) 045. [4] CMS Collaboration, JINST, 3 (2008) S08004.
[5] CMS Collaboration, Measurements of properties of the Higgs boson and search for an
additional resonance in the four-lepton final state at√s = 13 TeV, CMS-PAS-16-033 (2016).
[6] CMS Collaboration, Updated measurements of Higgs boson production in the di-photon
decay channel at√s = 13 TeV in pp collisions at CMS, CMS-PAS-16-020 (2016).
[7] CMS Collaboration, Search for associated production of Higgs bosons and top quarks in
multilepton final states at√s = 13 TeV, CMS-PAS-16-022 (2016).
[8] CMS Collaboration, Search for t¯tH production in the H→ b¯b decay channel with 2016 pp
collision data at√s = 13 TeV, CMS-PAS-16-022 (2016).
[9] CMS Collaboration, Updates on Projections of Physics Reach with the Upgraded CMS
