Search for b b decay of Higgs associated with a vector boson at ATLAS Lei Zhang on behalf of ATLAS collaboration International Symposium on Higgs Physics, IHEP, Beijing, 12-16, Aug. 2013 1
Introduction Higgs→b b is important for directly testing the Higgs mechanism in the • quark sector. • Due to the large multi-jet background at LHC, the inclusive search of H- >b b is almost impossible. • The Higgs associated production with W/Z boson is one of the most promising channels for H->b b search. • Three distinct channels, i.g. 0 lepton, 1 lepton and 2 lepton, have been explored mainly aiming to Z->vv, W->lv and Z->ll. 2
Introduction Higgs→b b is important for directly testing the Higgs mechanism in the • quark sector. • Due to the large multi-jet background at LHC, the inclusive search of H- >b b is almost impossible. • The Higgs associated production with W/Z boson is one of the most promising channels for H->b b search. • Three distinct channels, i.g. 0 lepton, 1 lepton and 2 lepton, have been explored mainly aiming to Z->vv, W->lv and Z->ll. HCP2012: ( ATLAS‐CONF‐2012‐161 ) Using combined 7TeV and partial 8TeV datasets, the observed (expected) limit is 1.8 (1.9) times the Standard Model prediction. 3
Introduction Higgs→b b is important for directly testing the Higgs mechanism in the • quark sector. • Due to the large multi-jet background at LHC, the inclusive search of H- >b b is almost impossible. • The Higgs associated production with W/Z boson is one of the most promising channels for H->b b search. • Three distinct channels, i.g. 0 lepton, 1 lepton and 2 lepton, have been explored mainly aiming to Z->vv, W->lv and Z->ll. HCP2012: ( ATLAS‐CONF‐2012‐161 ) Using combined 7TeV and partial 8TeV datasets, the observed (expected) limit is 1.8 (1.9) times the Standard Model prediction. In this talk, the latest ATLAS full run-I data result ( ATLAS‐CONF - 2013‐079) will be presented. 4
Event Selection • Common selections : – At least two jets with P T 1 >45GeV, P T 2 (or P T 3 )>20GeV and | η |<2.5 – Δ R(jet, jet) cut has been optimized as a function of P T V 1 lepton: 0 lepton: 2 lepton: – miss trigger 1-Lepton trigger+E T – – miss trigger E t 1 and 2 -lepton trigger – Exactly 1 lepton – – Veto leptons Exactly 2 leptons – M T W <120GeV and E T miss >25GeV – – E t miss (P T Z )>120 GeV 81GeV < M ll < 99GeV QCD rejection cuts: – QCD rejection cuts: E T miss <60GeV – W <160, M T W >40GeV P T – – – miss , jets) > 1.5 W >200, E T miss >50GeV Δφ (E T P T QCD negligible – miss , P T miss ) < π /2 Δφ E T – miss , bb) > 2.8 Δφ E T Enter 1-lepton analysis if one Enter 0-lepton channel if the charged lepton is not identified charged lepton is not identified 5
Background composition: post fit signal region (SR) 2-tag region plots Orange: 𝐮𝐮 2 jets Dark Orange: Wt channel Yellow: s and t channel Grean: W+jets Blue: Z+jets The heavier coler, the 3 jets heavier flavor jets. Pink : Multi-jet 0-lepton: (mixture) Control region(CR) following Z+jets , top and W+jets. the same color convention. 6
Background composition: post fit signal region (SR) 2-tag region plots 2 jets 3 jets 0-lepton: (mixture) 1-lepton: Z+jets , top and W+jets. top and W+jets. W bin. Some MJ at low P T 7
Background composition: post fit signal region (SR) 2-tag region plots 2 jets 3 jets 0-lepton: (mixture) 1-lepton: 2-lepton: Z+jets , top and W+jets. top and W+jets. Z+jets W bin. Z regions. Some MJ at low P T some top at low P T 8
Background modeling (I) Important corrections: tt sample (POWHEG+PYTHIA) : • – Top p T correction applied at the level of generated top quarks. • V+jets sample (LO SHERPA) : – Δφ (jet, jet) correction applied due to NLO effect (arXiv: 1207.5030v1) . – The P T V modeling greatly improved after correction. The most sensitive region With Δφ corr. No Δφ corr. 9
Background modeling (II) Multijet: (Data driven method) • Source: 0 lepton(mis-measured jets); 1 / 2 lepton (the jets faking leptons ) • The amount: 0 lepton ( 1% ); 1 lepton (15% - <1%); 2-lepton ( negligible). Non-multijet backgrounds: (M bb shape from MC) • Fixed to MC: Diboson, single-top, V+light-jets Float in fit: V+cl, V+bl, V+bb/cc, and tt • Inputs to Global fit 2jets, 1-tags 3jets, 1-tags 2jets, 2-tags 3jets, 2-tags Top e μ V bins 3 P T 0-lepton CR CR SR SR - V bins 5 P T 1-lepton CR CR SR SR - V bins 5 P T 2-lepton CR CR SR SR CR Control Region(CR) : Normalization Signal Region (SR) : Shape 10
Global fit Model: the idea • Adjust the normalization by simultaneous fit – Normalization floated: V+cl, V+bl/bb/cc(HF) and 𝐮𝐮 – Background scale factors have been correlated among regions. Illustrative plots 2lep,2tag e μ , 2-tag 2lep. 1tag 1lep,2tag 0lep,2tag 1lep. 1tag 11
Global fit Model: the idea • Adjust the normalization by simultaneous fit – Normalization floated: V+cl, V+bl/bb/cc(HF) and 𝐮𝐮 – Background scale factors have been correlated among regions. Illustrative plots 2lep,2tag e μ , 2-tag 2lep. 1tag 𝐮𝐮 1lep,2tag 0lep,2tag 1lep. 1tag 12
Global fit Model: the idea • Adjust the normalization by simultaneous fit – Normalization floated: V+cl, V+bl/bb/cc(HF) and 𝐮𝐮 – Background scale factors have been correlated among regions. Illustrative plots 2lep,2tag e μ , 2-tag 2lep. 1tag 𝐮𝐮 1lep,2tag 0lep,2tag 1lep. 1tag Z+cc/bb/bl 13
Global fit Model: the idea • Adjust the normalization by simultaneous fit – Normalization floated: V+cl, V+bl/bb/cc(HF) and 𝐮𝐮 – Background scale factors have been correlated among regions. Illustrative plots 2lep,2tag e μ , 2-tag 2lep. 1tag 𝐮𝐮 1lep,2tag 0lep,2tag 1lep. 1tag Z+cc/bb/bl W+cc/bb/bl 14
Global fit Model: the idea • Adjust the normalization by simultaneous fit – Normalization floated: V+cl, V+bl/bb/cc(HF) and 𝐮𝐮 – Background scale factors have been correlated among regions. Illustrative plots 2lep,2tag e μ , 2-tag 2lep. 1tag Z+cl 𝐮𝐮 1lep,2tag 0lep,2tag 1lep. 1tag Z+cc/bb/bl W+cl W+cc/bb/bl 15
Global fit Model: Constrain Systematic Uncertainties • Experimental: – JER/JES – B-tagging – Lepton ID – E T miss • Modeling: – M bb shape – P T V and Top pt – Jet multiplicity 16
Global fit Model: Constrain Systematic Uncertainties • Experimental: Migration between channels through – JER/JES lepton sysetmatics – B-tagging – Lepton ID For each – E T miss channel For each bin • Modeling: – M bb shape – P T V and Top pt For each – Jet multiplicity V bin P T 17
Fitted results: Diboson • VZ(bb): a similar signature with 5 times larger cross section than VH(bb) • Diboson fit: a validation of the Higgs analysis. • The obs. (exp.) significance of VZ is 4.8 (5.1) σ . μ VZ = 𝝉 𝐧𝐟𝐛𝐭 (𝐖𝐚) = 0.9 ± 0.2. • 𝝉 𝐓𝐍 (𝐖𝐚) 18
Fitted results: μ Higgs for each year and channel For m H = 125 GeV : • 7TeV: – 2 sigma deficit w.r.t. SM expectation. • 8TeV: – Consistent with both S+B and B only hypothesis in 1 σ . • Combined fit result: Results still dominated by statistical uncertainties. 19
Fitted results: Upper limit • 1 sigma excess in 8TeV data at m H =125GeV, as well as higher masses. • Deficit in 7TeV data which makes the combined results a small excess at m H =125 GeV. • At m H =125GeV, The observed (expected) upper limit at 95 C.L. is 1.4(1.3) times SM prediction. 20
Summary A search for H->b b has been performed via the Higgs associated • production with a vector boson by using full ATLAS run-I data. Diboson(VZ) result is consistent with the SM expectation, with 4.8 𝜏 • significance. For m H =125 GeV, VH →bb cross section combined fit results is: • • The observed (expected) 95% C.L. limit for m H =125 GeV is 1.4 (1.3) times the SM expectation. Respect to the previous analysis, a 35% improvement in significance on top of luminosity due to the optimization and reduced systematics. With run-I data, we are close to the critical region, and it will be very exciting to look the first bunch of data at 2015. 21
Th Than ank k yo you 22
backup 23
Trigger and Evt. selection Two lepton: One lepton: 1-Lepton miss trigger Zero lepton: E T 1-Lepton trigger miss trigger (for trigger + E T + 2-lepton trigger muon channel) 24
Xsec*Br and acceptance 25
0-lepton: data-MC plots 26
27
28
1-lepton: data-MC plots 29
30
31
2-lepton: data-MC plots 32
33
34
35
Background modeling (II): V+jets before Δφ (jet, jet) correction • V+jets: modeled by Leading order(LO) generator. • Mismodeling found in Δφ (jet, jet) and P T V distribution. • Interprated as Next Leading Order(NLO) effect --arxiv: 1207.5030v1 1 lepton, 2 jets, 0 tag V Δφ (jet, jet) P T High PTV region is the Δφ (jet, jet) correction has been applied most sensitive region. 36
Recommend
More recommend