Taikan Suehara et al., Tokyo, 12 Nov page 1 Taikan Suehara (Kyushu) A. Miyamoto (KEK), T. Tomita (Kyushu) Higgs recoil mass analysis: status and prospects in various channels

Taikan Suehara et al., Tokyo, 12 Nov page 2 Recoil in ILC g ZZH g EW not seen g EW Direct measurement of g ZZH with high precision cf. to LHC: NOT the “ratio” of Higgs couplings NOT suffered from QCD uncertainty

Taikan Suehara et al., Tokyo, 12 Nov page 3 Recoil mass Using 4-momentum conservation (not possible in LHC) Excellent mass resolution (lepton only: no jet ambiguity) ~ 30 MeV in ILC (1 st phase only) Usable as a cut to separate the main background ZZ (recoil mass peaked at 91.2 GeV) only looking 2f from Z ff

Taikan Suehara et al., Tokyo, 12 Nov page 4 global fit of higgs couplings Snowmass energy frontier reportModel independent limited by  250 GeV

Taikan Suehara et al., Tokyo, 12 Nov page 5 llh 250  3.0% (model independent) / 2.5% (semi model independent) Combining other modes –Higher energies: ILC500 –Jet recoil qqh 350/500 GeV 250 GeV Further improvements

Taikan Suehara et al., Tokyo, 12 Nov page 6 Cross section (  h) eL(0.8) pR(0.3) –250 GeV: 10 fb –500 GeV: 3.3 fb –Combining two may improve the resolution Recoil mass –Smeared in 500 GeV Background –Large t-channel diagram in 500 GeV Analysis –Almost the same as 250 GeV llh at 500 GeV

Taikan Suehara et al., Tokyo, 12 Nov page 7 Almost same as LoI Lepton ID Z mass (81.2 to GeV) di-lepton pT > 20 GeV recoil mass ( , ee) acoplanarity (  rad vetoed) pT balance |di-lepton pT – pT of the most energetic neutral particle| > 10 GeVCuts

Taikan Suehara et al., Tokyo, 12 Nov page 8 Same as LoI di-lepton pT cos  of di-lepton acolinearity Z mass TMVA used BDT & likelihood gives similar results Likelihood adoptedLikelihood

Taikan Suehara et al., Tokyo, 12 Nov page 9 mumu channel: cut table cuteehmmhooh2f4f6fo m>= e z mass ptdl> recoil acop ptbal like>

Taikan Suehara et al., Tokyo, 12 Nov page 10 BG fit (Gaus * 2 nd pol for mm, linear for ee) Fit will all paremeters free –GPET with 5 parameters Gaus (left side) Gaus + expo (right side) –background distribution from fit function to avoid large fluctuation due to the small stat Toy-MC – times –Poisson from data (sig), func (bg) –Fit with fixing shape parameters (mean and amplitude free) Recoil fit

Taikan Suehara et al., Tokyo, 12 Nov page 11 mumu: recoil fit 6.49%

Taikan Suehara et al., Tokyo, 12 Nov page 12 ee: brems recovery all neutral particle at cos  > 0.99 are added

Taikan Suehara et al., Tokyo, 12 Nov page 13 ee: cut table cuteehmmhooh2f4f6fo e>= e z mass ptdl> recoil acop ptbal l>

Taikan Suehara et al., Tokyo, 12 Nov page 14 ee: recoil fit 7.10%

Taikan Suehara et al., Tokyo, 12 Nov page 15 mmh: 6.49% eeh: 7.10% combined: 4.8% Summary of llh 500

Taikan Suehara et al., Tokyo, 12 Nov page 16 llh: only 6% of Z (e+  ) qqh: 70%  hopeful! h,  h: impossible to get recoil Jet clustering –efficiency can be different among higgs decay modes (6-jet, 4-jet, 2-jet) Jet energy resolution –Worse than leptons –Wider peak of recoil mass qqh recoil

Taikan Suehara et al., Tokyo, 12 Nov page 17 arXiv:

Taikan Suehara et al., Tokyo, 12 Nov page 18 qqh 350: W/Z separation sales point of PFA calorimeter A. Miyamoto

350 GeV result /09/04 ILD Analysis and Software meeting eR80.pR30 (300fb-1) eL80.pR30 (300fb-1) qqh f_h f_h f_sl f_l Higgs aa/ae/ea tt all-bkg S/sqrt(S+B) sqrt(S+B)/S S/N Selection statistics Recoil mass 123~133 GeV were counted. 150 fb-1 each to -80/+30 & +80/-30%  A. Miyamoto k t clustering w/ R=1.2 (may not be optimal) Basic cuts of Z mass etc.

500 GeV /09/04 ILD Analysis and Software meeting Event statistics for 500 fb-1, eL80.pR30 at 500 GeV ProcessNeventsFraction(%) Sqqh11113 B ffh f f f aa_2f ae/ea S/N= sqrt(S+N)/S=  Recoil mass Red: all background Black: qqh + background Hard to see recoil mass peak, but excess due to qqh could be seen

Summary A possibility to measure  ZH using using Z  qq mode was investigated. At 350 GeV, Higgs peak in jet recoil mass can be seen. Combining eL80/eR30 150fb -1 and eR80/eL30 150fb -1,  is expected. At 500 GeV, hard to see Higgs peak in jet recoil mass distribution. But from event excess in qqh like events,  is expected for 500 fb -1, eL80.pR30 beam polarization. Further improvement may be possible by more sophisticated analysis. A possibility to use Z  qq at 250 GeV should be investigated /09/04 ILD Analysis and Software meeting

Taikan Suehara et al., Tokyo, 12 Nov page GeV analysis (  ZH ~ maximum) –Durham y-fix clustering (should be optimized) –or dedicated Z  qq finder from PFOs Difference of efficiency in Higgs decay modes should be investigated (as systematic error study) Also used for detector optimization –performance with various jet energy resolution will be investigated qqh 250

Taikan Suehara et al., Tokyo, 12 Nov page 23 qqh 250: some plots ycut = ycut = black: all red: H  bb blue: H  WW* black: H  others njets

Taikan Suehara et al., Tokyo, 12 Nov page 24 qqh 250: some plots ycut = ycut = black: all red: H  bb blue: H  WW* black: H  others Z mass The combination with mass nearest to Z is selected bb  lower efficiency

Taikan Suehara et al., Tokyo, 12 Nov page 25 qqh 250: some plots ycut = ycut = black: all red: H  bb blue: H  WW* black: H  others recoil mass The combination with mass nearest to Z is selected cut on 81.2 < m Z < WW  many mis-jet pairings

Taikan Suehara et al., Tokyo, 12 Nov page 26 ycut fixed to Jet pairing done for 86 GeV (not 91.2 GeV) –To check the performance on W/Z separation Considering background of WW/ZZ  4q –suffered from mis-jet pairing/clustering Considering both e - L e + R and e - R e + L cases –P(e -,e + ) = (0.8,0.3) –Signal cross section larger in e - L e + R –WW suppressed in e - R e + L a bit more realistic

Taikan Suehara et al., Tokyo, 12 Nov page 27 di-jet mass ZH WW/ZZ e-Le+Re-Le+R not so powerful

Taikan Suehara et al., Tokyo, 12 Nov page 28 Recoil vs dijet mass ZH WW/ZZ di-jet mass recoil mass

Taikan Suehara et al., Tokyo, 12 Nov page 29 Recoil mass e-Le+Re-Le+R e-Re+Le-Re+L qqHWW/ZZsignificance e-Le+Re-Le+R 38,593991,50438  e-Re+Le-Re+L 26,032101,22773  Events with L = 250 fb -1, cut of m recoil > 120 GeV only VERY PRELIMINARY

Taikan Suehara et al., Tokyo, 12 Nov page 30 Remained issues improving results Optimization on jet clustering More cuts / MVAs Kinematic fit Remained issues degrading results Other backgrounds –WW/ZZ  qq + lepton/neutrinos –2f + ISR and others Efficiency depends on Higgs decay –Need study on systematic errors qqh 250 To do

Taikan Suehara et al., Tokyo, 12 Nov page 31 llh 250 GeV – 3.0/2.5% with 250 fb -1 e - L e + R Various channels for  ZH –llH 500 GeV – 4.8% with 500 fb -1 e - L e + R –qqH 350 GeV – 3.6% with fb -1 –qqH 500 GeV – 3.9% with 500 fb -1 e - L e + R –qqH 250 GeV (VERY PRELIMINARY) 2.6% with 250 fb -1 e - L e + R 1.4% with 250 fb -1 e - R e + L Study ongoing Systematic effects should be investigatedSummary