Searches for new physics at ATLAS using pair production of Higgs bosons. Jahred Adelman

Transcription

1 Searches for new physics at ATLAS using pair production of Higgs bosons Jahred Adelman Irvine NIU

2 Status of the Higgs boson back across the ocean ATLAS Preliminary m H = GeV Total uncertainty ±1σ on µ Phys. Rev. D 9, 1115 (14) H γγ arxiv: H ZZ* 4l arxiv: µ = µ = H WW* lνlν arxiv: W,Z H bb ATLAS-CONF H ττ +.3 µ = µ =.5-.4 So far, no major discrepancies from SM predictions +.4 µ = s = 7 TeV Ldt = fb s = 8 TeV Ldt =.3 fb Signal strength (µ) released Jahred Adelman Irvine

3 Status of the Higgs boson back across the ocean ATLAS-CONF-14-9,ZZ*,WW*, VBF+VH µ Higgs coupling to vector bosons m H Standard Model Best fit 68% CL 95% CL = 15.5 GeV ATLAS s = 7 TeV s = 8 TeV Preliminary Ldt = fb Ldt =.3 fb H H ZZ* 4l H WW* l l H µ,zz*,ww*, ggf+tth Higgs coupling to fermions Higgs boson-like resonance really looks like the SM Higgs boson Jahred Adelman Irvine 3

4 We are starting to probe the more difficult decays (h ττ) arxiv: Events / bin 4 3 Data Background (µ=1.4) Background (µ=) H (15) ττ (µ=1.4) H ττ ATLAS s = 8 TeV,.3 fb H (15) ττ (µ=1) 4.5σ evidence for h ττ decays 1 s = 7 TeV, 4.5 fb log (S / B) Jahred Adelman Irvine 4

5 And also the more rare production mechanisms (tth) ATLAS-CONF ATLAS Preliminary s=8 TeV, L dt=.3 fb Events / bin 5 4 ATLAS Preliminary Data 1 tth (µ =1.7) fit tth (µ =4.1) 95% excl. Bkgd Dilepton Lepton+jets Expected ± 1σ Expected ± σ 3 Combination Observed Expected ( µ=1) Data / Bkgd s=8 TeV L dt =.3 fb Comb. Single lepton and Dilepton tth (µ =1.7) + Bkgd. fit tth (µ =4.1) + Bkgd. 95% excl log (S/B) % CL limit on σ/σ SM at m H =15 GeV Limits of 4.1x SM expectation for tth production, just from bb channel Jahred Adelman Irvine 5

6 But of course... We are just being to probe and test this new particle Jahred Adelman Irvine 6

7 One long-term LHC goal arxiv: (Baglio et al), among many Observe the Higgs boson self-coupling, crucial to testing if the Higgs potential is the one predicted in the SM Jahred Adelman Irvine 7

8 SM Dihiggs production at the LHC SM hh production dominated by box diagram, not hh self-coupling, with destructive interference between the two Total SM hh cross section at 8 TeV ~9 fb not expected to be seen by us any time soon Jahred Adelman Irvine 8

9 What about extensions to the SM? arxiv: (Contino et al) (Kribs and Martin), (Baglio et al) among many Can enhance non-resonant hh production in many extensions to the SM tthh interactions, light colored scalars, if Higgs boson self-coupling were t h altered, or if top quark had nont standard Yukawa coupling t h Jahred Adelman Irvine 9

10 What about extensions to the SM? arxiv: (Baglio et al) Altered self-coupling can significantly increase hh production rates Jahred Adelman Irvine

11 Resonant production? arxiv:hep-ph/93 (Cheung), hep-ph/53173 (Djouadi), (Dolan et al), (Tang), (Kumar & Martin), among many Can enhance hh production resonantly as well Two Higgs doublet models, Randall-Sundrum gravitons, radions, stoponium,... Jahred Adelman Irvine 11

12 How to go about looking for Higgs bosons? ET = 66.8 GeV mᵧᵧ = 15.8 GeV ET = 56.9 GeV Jahred Adelman Irvine 1

13 Zooming in on a converted photon Jahred Adelman Irvine 13

14 The calorimeters in action Invariant mass of two most energetic jets = 4.1 TeV Jahred Adelman Irvine 14

15 The muon subsystems m4l = 14.3 GeV mµµ = 76.8 GeV mee = 45.7 GeV Jahred Adelman Irvine 15

16 Looking for hh bbγγ. Why? At known mh, h bb has highest Higgs BR (.57) h γγ has high efficiency and good mass resolution Can perform full mass reconstruction h ττ and h WW have poor mass resolution vs γγ h ZZ few events after require leptonic decays Sensitive to lower mass resonances and also the region testing hhh vertex M H Jahred Adelman Irvine 16 Higgs BR + Total Uncert cc µµ bb gg arxiv: Z WW ZZ LHC HIGGS XS WG 13

17 Looking for hh bbγγ arxiv: Start with common ATLAS h γγ selection Loose diphoton trigger nearly % efficient for offline cuts ET >.35(.5)mᵧᵧ for leading (subleading) photon η <.37 excluding 1.37 < η < 1.56 Higgs BR + Total Uncert cc µµ bb gg arxiv: Z WW ZZ M H LHC HIGGS XS WG 13 Jahred Adelman Irvine 17

18 Photons for Higgs analyses ATLAS-COM-PHYS-1-36, ATLAS-COM-PHYS-1-83 Use tight photon ID. Can have unconverted, 1-track converted or -track converted photons, all isolated ΣpT (tracks with pt > 1 GeV) in cone dr <. from photon <.6 GeV ΣET (calorimeter) in cone dr <.4 from photon < 6 GeV, corrected for γ energy leakage and pileup Fraction of photon candidates Unconverted photons Converted photons Single track conversions Double track conversions ATLAS Preliminary Data 1, s = 8 TeV L dt = 3.3 fb Isolation energy Data (Z e + e - ), L dt = 3 fb Simulation (shifted by MeV) Large gap 8 BCID gap ATLAS Preliminary Average interactions per bunch crossing - 3 Bunch crossing ID Jahred Adelman Irvine 18

19 Jets and b-tagging ATLAS-CONF-14-4 Require two anti-kt R=.4 jets with η <.5 Perform b-tagging using neural network tagger at 7% efficiency for b-jets in simulated ttbar events Rejection factor 13x (4x) for light quark (charm) jets Calibrate b-tag scale factors using dilepton ttbar events b-jet efficiency ATLAS tt PDF (MC) tt PDF (Data) 3 4 Preliminary L dt =.3 fb s = 8 TeV MV1, b = 7% Jet p T Jahred Adelman Irvine 19

20 Jets and bb invariant mass Leading jet pt > 55 GeV, subleading pt > 35 GeV after adding in 4-vectors of any muons with pt > 4 GeV with dr <.4 to jet Require 95 < mbb < 135, 75% efficiency for hh Asymmetric cut optimized in simulation and largely due to energies losses from escaping neutrinos Fractional JES uncertainty R =.4, LCW+JES + in situ anti-k t Data 1, =. 3 4 s = 8 TeV correction Total uncertainty Absolute in situ JES Relative in situ JES Flav. composition, inclusive jets Flav. response, inclusive jets Pileup, average 1 conditions ATLAS Preliminary 3 jet p T Jahred Adelman Irvine Fractional JES uncertainty R =.4, LCW+JES + in situ anti-k t Data 1, p jet T = 4 GeV s = 8 TeV correction Total uncertainty Absolute in situ JES Relative in situ JES Flav. composition, inclusive jets Flav. response, inclusive jets Pileup, average 1 conditions ATLAS Preliminary

21 Strategy for hh bbγγ Start by looking for non-resonant production of hh bbγγ Signal region is two photons with invariant mass consistent with mh + two b-tagged jets with mass loosely consistent with mh Cuts optimized for discovery while trying to maintain as simple a selection as possible at the same time Jahred Adelman Irvine 1

22 Blinded Continuum background estimation for hh bbγγ arxiv: All backgrounds without a Higgs boson estimated using m(γγ) sideband, which is fit to an exponential and extrapolated to signal region Normalization comes from sidebands Slope of the exponential constrained by fitting events with < b-tags Events /.5 GeV Events /.5 GeV ATLAS Ldt = fb, s = 8 TeV Signal Region Data Fitted Signal + Bkds Single Higgs Boson + Bkd Continuum Background < b-tag Control Region m γγ m γγ Jahred Adelman Irvine

23 Blinded Continuum background estimation for hh bbγγ Take advantage of good diphoton mass resolution ~1.6 GeV Note the normalization uncertainty of ~3-35% that cannot be avoided Events /.5 GeV 8 6 ATLAS Ldt = fb, s = 8 TeV Signal Region Data Fitted Signal + Bkds Single Higgs Boson + Bkd Continuum Background 4 Events /.5 GeV < b-tag Control Region m γγ Jahred Adelman Irvine 3

24 Higgs background estimation for hh bbγγ Smaller backgrounds with single Higgs boson estimated from simulation W/Z/tt + Higgs from Pythia8 WH xsec at NLO with EW corrections ZH xsec at NNLO with EW corrections tth xsec at NLO σ(pp H+X) [pb] 1 pp H (NNLO+NNLL QCD + NLO EW) pp qqh (NNLO QCD + NLO EW) pp WH (NNLO QCD + NLO EW) pp ZH (NNLO QCD +NLO EW) arxiv: (LHC XS WG) s= 8 TeV LHC HIGGS XS WG 1 pp tth (NLO QCD) M H Jahred Adelman Irvine 4

25 Higgs background estimation for hh bbγγ gg and VV fusion with Powheg-Box ggf xsec at NNLO+leading-log resummation and EW corrections VV fusion at NLO+EW and approximate NNLO corrections bbh estimated to be negligible with our pt and mass cuts σ(pp H+X) [pb] 1 pp H (NNLO+NNLL QCD + NLO EW) pp qqh (NNLO QCD + NLO EW) pp WH (NNLO QCD + NLO EW) pp ZH (NNLO QCD +NLO EW) s= 8 TeV LHC HIGGS XS WG 1 pp tth (NLO QCD) M H Jahred Adelman Irvine 5

26 Signals for hh bbγγ SM pair production of hh produced with Madgraph5+Pythia8, including interference between diagrams Resonant hh production modeled with a gluoninitiated spin- resonant state in a narrow-width approximation (NWA) qq-initiated NWA signals give very similar kinematics and efficiencies Radions would be in the NWA Gravitons would not be in the NWA Most interesting regions (mh < mh < mt) of HDM phase space are in the NWA Jahred Adelman Irvine 6

27 Systematic uncertainties for hh bbγγ All small compared to statistical uncertainties Systematic uncertainty Non-Resonance Analysis Single h Bkgd hh Signal Continuum Trigger [%].5 Luminosity [%].8 Photon Mass Shape Jets Theory Identification [%].4 Isolation [%] Resolution [%] Resolution: 13 Position Value: +.5/-.6 GeV m Continuum Shape [%] 11 m bb :Statistical[%] m bb : jj vs bb [%] m bb :FitModel[%] b-tagging [%] Energy Scale [%] b-jet Energy Scale [%].6.3 Energy Resolution [%] PDF+Scale [%] 8.4 Single h+hf [%] 14 Fit sidebands to -tag data, 1-tag, data with nonisolated photons, and using flat function (largest=11%) % uncertainty on gg and VBF due to HF content Jahred Adelman Irvine 7

28 Backgrounds for hh bbγγ Process Fraction of total ggh 11% qqh % WH 1% ZH 17% t th 69% Total.17 ±.4 Events Compares with.4 hh events and 1.3 events from continuum backgrounds in ±σ(mᵧᵧ) Continuum split evenly between γγjj and γjjj j can be b, c or light ttbar ~ % of the total Jahred Adelman Irvine 8

29 Results for non-resonance hh bbγγ search Events /.5 GeV Events /.5 GeV ATLAS Signal Region Data Ldt = fb, s = 8 TeV Fitted Signal + Bkds Single Higgs Boson + Bkd Continuum Background < b-tag Control Region m γγ Unbinned S+B fit 1.5 background events expected 5 events observed.4σ from backgroundonly hypothesis 95% CL upper limit on hh production of. pb (expected 1. pb) Jahred Adelman Irvine 9

30 Moving to resonant production Continue with previous analysis but add additional feature of resonance in 4-object invariant mass Jahred Adelman Irvine 3

31 Searching for X hh bbγγ Useful to improve 4-object invariant mass resolution as much as possible to reject background while maintaing signal efficiency Once we select objects, require mbb to give back 15 GeV (by scaling the combined bb 4-vector) Fraction /.5 GeV ATLAS Simulation s = 8 TeV mx =6 GeV m =3 GeV m m X X X =35 GeV =5 GeV Constrained Constrained Constrained Constrained Improves mass resolution by 3-6% m γγbb Jahred Adelman Irvine 31

32 How to use rescaled mass Require mγγbb to be within window around resonance mass mh with 95% signal efficiency Window varies from 17 GeV (mx = 6 GeV) to 6 GeV (mx = 5 GeV) Fraction /.5 GeV ATLAS Simulation s = 8 TeV mx =6 GeV m =3 GeV m m X X X =35 GeV =5 GeV Constrained Constrained Constrained Constrained Jahred Adelman Irvine 3 m γγbb

33 Efficiency for continuum to pass this cut? Measure efficiency of continuum to pass this cut using events with < b-tags, fit with a priori Landau function For mx = low (6 GeV) and high (5 GeV) mh, efficiency for continuum < 8% For mx = 3 GeV, 18% of continuum cut Events / GeV 1 ATLAS Ldt = fb, s = 8 TeV Signal Region Data < b-tag Control Region Data Landau Fit Control Region Fit 3 4 Single 5Higgs Boson m X =3 GeV, σ X BR hh =1 pb Constrained m γ γ jj Jahred Adelman Irvine 33

34 How to perform the resonance search? Cannot fit sidebands after resonance selection (likely no events left) Instead, perform cut-and-count analysis around ±σ of mh (in mγγ) and in a given window of mγγbb Jahred Adelman Irvine 34

35 The resonance search Jahred Adelman Irvine 35

36 Systematic uncertainties in resonance search All very small compared to statistical uncertainties Resonance Analysis SM h + hh Bkgd H! hh Signal Continuum Migration: 1.6 Migration: 1.7% /5 14 Use simulation to evaluate differences in shape between γγbb and γγjj masses Use alternate fit functions to Landau distribution Jahred Adelman Irvine 36

37 mγγbb mass in resonance search Events / 5 GeV 1 - ATLAS Ldt = fb, s = 8 TeV Signal Region Data Control Region Fit Single Higgs Boson m X =3 GeV, σ X BR hh =1 pb Same 5 events in SR as before Events / GeV 1 < b-tag Control Region Data Landau Fit Constrained m γ γ jj Jahred Adelman Irvine 37

38 Limits in resonance search BR(X hh) [pb] σ X ATLAS Ldt = fb at s = 8 TeV Observed 95% CL Limit Expected Limit ±1σ Expected Limit ±σ Type I HDM: tanβ=1, cos(β-α)= Jahred Adelman Irvine m X 38

39 Event displays mγγbb = 89.9 GeV mᵧᵧ = 15.1 GeV Jahred Adelman Jets Photons Muons Irvine 39

40 Event displays mγγbb = GeV mᵧᵧ = 15. GeV Jahred Adelman Jets Photons Muons Irvine 4

41 p-values in resonance search Local p 1 - ATLAS Ldt = fb at s = 8 TeV Observed p σ 1σ σ -3 3σ Jahred Adelman Global p-value =.1σ Irvine m X 41

42 Switching topics again Jahred Adelman Irvine 4

43 Searching for X hh bbbb Search for resonant pair production of Higgs bosons via X hh bbbb Dominant background is multi-jet background, subleading background is ttbar production Jahred Adelman Irvine 43

44 Searching for X hh bbbb ATLAS-CONF b analysis much more sensitive at high mx, where the backgrounds can be kept to a manageable rate Look for mx starting at 5 GeV Larger signals, but multijet backgrounds require data-driven methods to trust predictions Benchmark signal model: Randall-Sundrum graviton with a warped extra dimension k/mpl set to unity Resonance smaller than 4j mass resolution Modeled with Madgraph interfaced to Pythia8 Jahred Adelman Irvine 44

45 4b triggers Signal Trigger Efficiency ATLAS Simulation Preliminary EF_b45_medium_4j45_a4tchad_LFS EF_b35_loose_j145_j35_a4tchad EF_b45_medium_j145_j45_a4tchad_ht5 EF_4j8_a4tchad_LFS EF_j36_a4tchad m G* Combination of 5 non-prescaled triggers >99.5% efficiency with respect to offline selection for mass range of interest Jahred Adelman Irvine 45

46 4b event selection Require 4 b-tagged jets with pt > 4 GeV Look for dijet systems with pt(jj) > GeV and dr(j,j) < 1.5 Look for extra jets with pt > 3 GeV and dr(jj) < 1. to form ttbar system, and calculate:.1mw.1mt mt from 3-jet mass, mw mass of extra jet and b-jet with lowest b-tag probability Reject events if Xtt < 3. (keeps 9% signal, rejects 6% ttbar background) Jahred Adelman Irvine 46

47 4b event selection Form XHH from pairs of dijets:.1mlead m ~ lead = 14 GeV, m ~ subl = 115 GeV, optimized in simulation.1 Require XHH < 1.6 to be.1.8 in signal region.1msubl m G* Jahred Adelman Irvine 47 A x.14 ATLAS Simulation Preliminary 4 b-tagged jets dijets tt veto Signal Region (HH).6.4.

48 4b multijet estimate in tag data subl m dijet Sideband region ZZ 5 ATLAS Preliminary s = 8 TeV: Ldt = 19.5 fb HZ/ZH Control region (6,16) lead m dijet Signal region Events / 4 GeV ZZ, HZ, ZH regions defined in same way as HH but with masses set for consistency with different final states Jahred Adelman Irvine 48

49 4b multijet estimate subl m dijet Sideband region Control region ATLAS Preliminary s = 8 TeV: Ldt = 19.5 fb Signal region Events / 4 GeV Two-tag data used to estimate multijet background in 4tag data ZZ HZ/ZH lead m dijet µ=.64 ±. Jahred Adelman Irvine 49

50 Kinematics of estimate Data / Bkgd Events / 5 GeV Data / Bkgd Events / 5 GeV 1 Sideband Region Data Multijet tt m 4j 1 Sideband Region Data Multijet tt ATLAS Preliminary s = 8 TeV: Ldt = 19.5 fb ATLAS Preliminary s = 8 TeV: Ldt = 19.5 fb m 4j Before After Refine kinematics by reweighting from sideband events in leading dijet pt, dr(j,j) in subleading dijet, dr(dijets) Jahred Adelman Irvine 5

51 4b check in control region Residual normalization and shape differences taken as uncertainty Jahred Adelman Irvine 51

52 4b ttbar estimate ttbar contribution estimated from events where at least one dijet system fails Xtt Extrapolation to signal region from semi-leptonic ttbar events in data Large statistical uncertainties, but ttbar background is small compared to multijet background Shape of ttbar events taken from simulation Jahred Adelman Irvine 5

53 4b systematic uncertainties Relative change in expected limit Source of Uncertainty B-Tagging JER JES Luminosity Multijet Normalisation Multijet Shape tt Normalisation tt Shape s ATLAS Preliminary = 8 TeV: Ldt = 19.5 fb B-tagging scale factor uncertainties for jets with pt > 3 GeV large (not enough data - estimated purely from simulation) Jahred Adelman Irvine 53 m G*

54 4b results subl m dijet 3 5 ATLAS Preliminary s = 8 TeV: Ldt = 19.5 fb 3 5 Events / 4 GeV lead m dijet 15 5 Jahred Adelman Irvine 54

55 4b results σ(pp G*) x BR(G* HH bbbb) [fb] 1 ATLAS Preliminary Expected Limit (95% CL) Expected ± 1σ Expected ± σ Observed Limit (95% CL) RS Graviton, k/m Planck = 1. s = 8 TeV: Ldt = 19.5 fb m G* Exclude gravitons at 95% CL with mass 59-7 GeV (expected GeV) Jahred Adelman Irvine 55

56 Event display mbb(1) = 115 GeV mbb() = 11 GeV m4b = 834 GeV Jahred Adelman Irvine 56

57 Last thoughts before concluding arxiv: (Dolan et al) Can we use kinematics to separate out box and triangle diagrams?! arxiv: (Slawinska et al) Jahred Adelman Irvine 57

58 Last thoughts before concluding arxiv: (Baglio et al) S p =1.3 B vs S p =6.5 B ATL-PHYS-PUB Jahred Adelman Irvine 58

59 Conclusions Key LHC goal to explore the Higgs boson potential and study the Higgs self-coupling For the near future, we are searching for new physics In the long-term, we hope to begin to make measurements (hopefully mostly of triangle!) A difficult path, but also a fun one Jahred Adelman Irvine 59

60 Conclusions Look for Run 1 γγww, ττbb and boosted-4b analyses (out very soon!) and a combination, and γγbb searches early in Run Jahred Adelman Irvine 6

61 Thank you! Jahred Adelman Irvine 61

62 Backup Jahred Adelman Irvine 6

63 Photon ID Jahred Adelman Irvine 63 (tight) ID <.6 unconverted iso E T < 4 GeV ATLAS Preliminary s = 8 TeV Ldt =.3 fb Electron extrapolation Matrix method Z ll (tight) ID < 1.37 unconverted iso E T < 4 GeV ATLAS Preliminary s = 8 TeV Ldt =.3 fb Electron extrapolation Matrix method Z ll < ID > - ID E T < ID > - ID E T (tight) ID < < 1.81 unconverted < 4 GeV iso E T ATLAS Preliminary s = 8 TeV Ldt =.3 fb Electron extrapolation Matrix method Z ll (tight) ID <.37 unconverted iso E T < 4 GeV ATLAS Preliminary s = 8 TeV Ldt =.3 fb Electron extrapolation Matrix method Z ll < ID > - ID E T < ID > - ID E T

64 Photon ID Jahred Adelman Irvine 64 (tight) ID <.6 converted iso E T < 4 GeV ATLAS Preliminary s = 8 TeV Ldt =.3 fb Electron extrapolation Matrix method Z ll (tight) ID < 1.37 converted iso E T < 4 GeV ATLAS Preliminary s = 8 TeV Ldt =.3 fb Electron extrapolation Matrix method Z ll < ID > - ID E T < ID > - ID E T (tight) ID < 1.81 converted iso E T < 4 GeV ATLAS Preliminary s = 8 TeV Ldt =.3 fb Electron extrapolation Matrix method Z ll (tight) ID <.37 converted iso E T < 4 GeV ATLAS Preliminary s = 8 TeV Ldt =.3 fb Electron extrapolation Matrix method Z ll < ID > - ID E T < ID > - ID E T

65 Photon ID Jahred Adelman Irvine 65 (tight) combined ID error on ID ATLAS Preliminary 3 4 s = 8 TeV, Ldt =.3 fb <.6 unconverted iso E T < 4 GeV E T (tight) combined ID error on ID ATLAS Preliminary 3 4 s = 8 TeV, Ldt =.3 fb.6 < 1.37 unconverted iso E T < 4 GeV E T (tight) combined ID error on ID ATLAS Preliminary 3 4 s = 8 TeV, Ldt =.3 fb 1.5 < 1.81 unconverted iso E T < 4 GeV E T (tight) combined ID error on ID ATLAS Preliminary 3 4 s = 8 TeV, Ldt =.3 fb 1.81 <.37 unconverted iso E T < 4 GeV E T

66 Photon ID Jahred Adelman Irvine 66 (tight) combined ID error on ID ATLAS Preliminary 3 4 s = 8 TeV, Ldt =.3 fb <.6 converted iso E T < 4 GeV E T (tight) combined ID error on ID ATLAS Preliminary 3 4 s = 8 TeV, Ldt =.3 fb.6 < 1.37 converted iso E T < 4 GeV E T (tight) combined ID error on ID ATLAS Preliminary 3 4 s = 8 TeV, Ldt =.3 fb 1.5 < 1.81 converted iso E T < 4 GeV E T (tight) combined ID error on ID ATLAS Preliminary 3 4 s = 8 TeV, Ldt =.3 fb 1.81 <.37 converted iso E T < 4 GeV E T

67 hh xsections in SM Jahred Adelman Irvine 67

68 yybb events Jahred Adelman Irvine 68

69 HDM arxiv: hep-ph/ Extend the Higgs sector to a second Higgs doublet Type 1: All fermions couple to one doublet Type : Up-type quarks couple to one doublet, down-type quarks and leptons to the second doublet Type 3: Quarks couple to one doublet, leptons to second doublet Type 4: Up-type quarks and leptons couple to one doublet, down-type quarks to the second doublet tan β = ratio of vev of both doublets α determines mixing between two neutral scalars Jahred Adelman Irvine

70 Higgs boson production Jahred Adelman Irvine 7 Analogous diagrams for hh production, too!

71 4b background Jahred Adelman Irvine 71 Events / 5 GeV Signal Region Multijet tt ATLAS Preliminary s = 8 TeV: Ldt = 19.5 fb m 4j

72 4b multijet estimate Events / 5 GeV 1 Sideband Region Data Multijet tt 8 Events / 5 GeV 1 Sideband Region Data Multijet tt ATLAS Preliminary s = 8 TeV: Ldt = 19.5 fb 6 4 ATLAS Preliminary s = 8 TeV: Ldt = 19.5 fb Data / Bkgd m 4j Data / Bkgd m 4j Before reweighting After reweighting Jahred Adelman Irvine 7

73 4b multijet estimate in tag data subl m dijet Sideband region ZZ 5 ATLAS Preliminary s = 8 TeV: Ldt = 19.5 fb HZ/ZH Control region (6,16) lead m dijet Signal region Events / 4 GeV ZZ region defined in same way as HH but with masses set to 86 and 93 GeV, and XZZ < 1.5 HZ (ZH) region has mass windows set to 14/86 (93/115) GeV and X < 1.6 Two-tag data used to estimate multijet background in 4tag data Jahred Adelman Irvine 73

74 Photons for Higgs analyses ATLAS-COM-PHYS Efficiency extrapolated from Z ee and measured in Z llγ + matrix method using track isolation sidebands (tight) ID < ID > - ID <.6 converted < 4 GeV iso E T ATLAS Converted central photons Preliminary 3 4 s = 8 TeV Ldt =.3 fb Electron extrapolation Matrix method Z ll E T (tight) ID < ID > - ID <.6 unconverted < 4 GeV iso E T Unconverted central photons ATLAS Preliminary 3 4 s = 8 TeV Ldt =.3 fb Electron extrapolation Matrix method Z ll E T Jahred Adelman Irvine 74

75 Comparison to CMS CMS-PAS-HIG3-3 Those who pay attention to these things may have noticed that CMS also has a recent CONF note searching for hh resonances in the γγbb channel Jahred Adelman Irvine 75

76 Comparing ATLAS and CMS ATLAS CMS-PAS-HIG3-3 CMS Jet pt 55/35 GeV 5 GeV Tag requirement tag Separate 1tag and tag regions for signal mjj range 9535 GeV 8555 GeV mjj method 4-vector scaling Kinematic fit Resonance limit method Counting experiment Sideband fit Non-resonance limit Yes No Signal at 3 GeV Background at 3 GeV Limit at 3 GeV CMS ~5% larger in -tag channel CMS ~4% larger in -tag channel CMS ~5% better (expected) Jahred Adelman Irvine 76