Developments on SM Higgs-boson cross sections and branching ratios

Transcription

1 Developments on SM Higgs-boson cross sections and branching ratios Laura Reina Higgs Hunting, Orsay, France

2 Outline Living through the most exciting time for Higgs physics: crucial to have access to the best theoretical predictions for SM Higgs-boson cross sections and branching ratios. foundation of the Tevatron exclusion limits; spirit of the LHC Higgs Cross Section Working Group ( State of the art of inclusive results for a SM Higgs boson ( see Grazzini s talk for in depth analysis of main channels). Towards exclusive results: goals, challenges, highlights from existing studies. Results on branching ratios.

3 Motivation: with s = 7 TeV and a few fb... Combining mainly H W + W, H ZZ, H γγ, ATLAS and CMS indicate that, if no signal, the SM Higgs can be excluded up to 5 GeV; a 5σ significance for a SM Higgs in the 4 7 GeV mass range. gain by adding new channels and optimizing cuts ( see Metha s talk). Significance, σ fb - fb - 5fb s=7 TeV s=8 TeV s=7 TeV s=8 TeV fb - ATLAS Preliminary (Simulation) m H [GeV] Need adequate theoretical input and careful matching between th. and exp.

4 Inclusive SM Higgs-boson production cross sections (LHC Higgs Cross Section WG, ) (arxiv:.593 CERN Yellow Report) MH [GeV] pp tth (NLO QCD) pp WH (NNLO QCD + NLO EW) pp ZH (NNLO QCD +NLO EW) σ(pp H+X) [pb] pp qqh (NNLO QCD + NLO EW) pp H (NNLO+NNLL QCD + NLO EW) s= 7 TeV MH [GeV] pp ZH (NNLO QCD +NLO EW) pp WH (NNLO QCD + NLO EW) pp qqh (NNLO QCD + NLO EW) LHC HIGGS XS WG σ(pp H+X) [pb] pp H (NNLO+NNLL QCD + NLO EW) s= 4 TeV Implemented a coherent Higgs precision program: all orders of calculated higher orders corrections included (tested with all existing calculations); common recipe for renormalization+factorization scale dependence; PDF and αs errors following PDF4LHC prescription ( see de Florian s talk); all other parametric errors included; theory errors combined according to common recipe. pp tth (NLO QCD) LHC HIGGS XS WG

5 For s = 7 TeV (from S. Dittmaier s talk, BNL, May ) Uncertainties NLO/NNLO/NNLO+ M H scale PDF4LHC QCD EW ggf < 5 GeV 6-% 8-% > % 5% VBF < 5 GeV % -7% 5% 5% W H < 3 GeV % 3-4% 3% 5-% ZH < 3 GeV -% 3-4% 4% 5% t th < 3 GeV % 9% 5%? For s = 4 TeV Uncertainties NLO/NNLO/NNLO+ M H scale PDF4LHC QCD EW ggf < 5 GeV 6-4% 7% > % 5% VBF < 5 GeV % 3-4% 5% 5% W H < 3 GeV % 3-4% 3% 5-% ZH < 3 GeV -4% 3-4% 45% 5% t th < 3 GeV % 9% 5-%?

6 Based on several contributions: Higgs process σ NLO,NNLO,NNLL,EW gg H q q (W,Z)H q q q qh q q,gg t th S.Dawson, NPB 359 (99), A.Djouadi, M.Spira, P.Zerwas, PLB 64 (99) C.J.Glosser et al., JHEP (); V.Ravindran et al., NPB 634 () D. de Florian et al., PRL 8 (999) R.Harlander, W.Kilgore, PRL 88 () (NNLO) C.Anastasiou, K.Melnikov, NPB 646 () (NNLO) V.Ravindran et al., NPB 665 (3) (NNLO) S.Catani et al. JHEP 37 (3) (NNLL) G.Bozzi et al., PLB 564 (3), NPB 737 (6) (NNLL) C.Anastasiou, R.Boughezal, F.Petriello, JHEP (8) (QCD+EW) T.Han, S.Willenbrock, PLB 73 (99) M.L.Ciccolini, S.Dittmaier, and M.Krämer (3) (EW) O.Brien, A.Djouadi, R.Harlander, PLB 579 (4) (NNLO) T.Han, G.Valencia, S.Willenbrock, PRL 69 (99) T.Figy, C.Oleari, D.Zeppenfeld, PRD 68 (3) M.L.Ciccolini, A.Denner,S.Dittmaier (8) (QCD+EW) P.Bolzoni, F.Maltoni, S.O.Moch, and M.Zaro () (NNLO) W.Beenakker et al., PRL 87 (), NPB 653 (3) S.Dawson et al., PRL 87 (), PRD 65 (), PRD 67,68 (3)

7 Towards exclusive studies: including decays, cuts, jet vetos, backgrounds,... Provide distributions from NLO/NNLO/NNLL calculations. Study the impact of higher order corrections in the presence of cuts, jet vetos, etc. If cuts imposed on decay products, need to include decays and estimate higher order corrections to the new process high multiplicity of final state makes calculation more involved (more and more NLO calculations coming on-line) narrow width approximations often excellent approximation (top, light Higgs) (Ex.: Melnikov, Schulze, arxiv:6.9, arxiv:.967) Interface with NLO Monte Carlo programs should be implemented and results compared: POWHEG. Backgrounds need to be calculated with comparable accuracy. Signal-background interference needs to be carefully adressed. More channel-specific issues...

8 Magnitude of higher order corrections varies significantly with signal selection cuts and vetoes. Ex.: (gg )H W + W l + νl ν [Anastasiou,Dissertori,Stöckli (7)] ( see also Grazzini s talk)

9 Main issues: Inclusive studies not indicative for exclusive predictions. Logarithmic dependence from extra scales (cuts/vetos) interferes with usual µ R and µ F -dependence: difficult to estimate overall theoretical uncertainty (very cut/veto-dependent). Need to question stability of perturbative prediction Need dedicated studies, for all channels/analyses: availability of NLO (and NNLO if needed) codes becomes mandatory. The exercise we are now completing for SM Higgs searches is a glorious application of the incredible progress in NLO calculations over the past few years.

10 NLO: challenges have largely been faced and enormous progress has been made ( see also Maltoni s talk) several independent codes based on traditional FD s approach several NLO processes collected and viable in MFCM [Campbell, Ellis] Enormous progress towards automation: Virtual corrections: new techniques based on unitarity methods and recursion relations BlackHat [Berger, Bern, Dixon, Febres Cordero, Forde, Ita, Kosower, Maitre] Rocket [Ellis, Giele, Kunszt, Melnikov, Zanderighi] HELAC+CutTools,Samurai [Bevilacqua, Czakon, van Harmeren, Papadopoulos, Pittau,Worek; Mastrolia, Ossola, Reiter, Tramontano] Real corrections: based on Catani-Seymour Dipole subtraction or FKS subtraction Sherpa [Gleisberg, Krauss] Madgraph (AutoDipole) [Hasegawa, Moch, Uwer] Madgraph (MadDipole) [Frederix, Gehrmann, Greiner] Madgraph (MadFKS) [Frederix,Frixione, Maltoni, Stelzer]

11 virtual+real: MadLoop+MadFKS [Hirschi, Frederix, Frixione, Garzelli, Maltoni, Pittau] interface to parton shower well advanced: POWHEG [Nason, Oleari, Alioli, Re] [Frixione, Webber, Nason, Frederix, Maltoni, Stelzer] [Frederix, Frixione, Hirschi, Maltoni, Pittau, Torrielli] Tools that we can now use for signal and ( high multiplicity) background. A choice of examples to follow...

12 W+jets Entries / GeV/c Data s = 7 TeV L dt = 35 pb - ATLAS Preliminary Dibosons top W/Z+jets Multi-jet Signal (x3) m H = 4 GeV/c H + j Search m WW [GeV/c ] jets) [pb] jet σ(w + N 4 3 Ldt=33 pb - W eν + jets Data, s=7 TeV ALPGEN SHERPA PYTHIA BLACKHAT-SHERPA MCFM ATLAS Preliminary Entries / GeV/c L dt = 35 pb s = 7 TeV - ATLAS Preliminary Data Dibosons top W/Z+jets Multi-jet Signal (x3) m H = 4 GeV/c H + j Search m WW [GeV/c ] Theory/Data Inclusive Jet Multiplicity, N Blackhat+Sherpa: W+3j, W+4j at NLO jet

13 W + b-jets: crucial background for W H production Two interesting signatures: W +b jets (m b ): Febres Cordero, L. R., Wackeroth, arxiv:hep-ph/66, arxiv:96.93 Badger, Campbell, Ellis, arxiv:.6647 Oleari, L. R., arxiv POWHEG Frederix, Frixione, Hirschi, Maltoni, Pittau, Torrielli amc@nlo W + jets with at least one b jet: Campbell, Ellis, Febres Cordero, Maltoni, L. R., Wackeroth, Willenbrock, arxiv:89.33 the CDF collaboration, arxiv:99.55, Campbell, Febres Cordero, L. R., arxiv:.336, arxiv:.954 the ATLAS collaboration, A. Messina s talk at EPS, Caola, Campbell, Febres Cordero, L. R., Wackeroth, arxiv:7.374

14 In a nutshell: One or two LO processes, depending on choice of 4FNS vs 5FNS: q b b b b q q + O(α s ) corrections q Correspondently, at NLO: W W. q q Wb b at tree level and one loop (m b ). q q Wb bg at tree level (m b ) 3. bq Wbq at tree level and one loop (m b = ) 4. bq Wbq g and bg Wbq q at tree level (m b = ) 5. gq Wb bq at tree level (m b ) avoiding double counting: b g q b q b q q b b(x,µ) = α s π ln µ m b x dy y P qg( x y) g(y,µ) W W +b jets: processes ++5 W + jets with at least one b jet: processes W

15 Comparison with CDF measurement: a puzzle? CDF Note 93 (arxiv:99.55): σ b jet (W +bjets) Br(W lν) =.74±.7(stat)±.4(syst) pb From our W +b calculation: [Neu, Thomson, Heinrich] [Campbell, Febres Cordero, L.R.] σ b jet (W +bjets) Br(W lν) =.±.4 pb For comparison: ALPGEN prediction:.78 pb PYTHIA prediction:. pb

16 Comparison with ATLAS [Golling, Messina, et al.]

17 Further development: Wb b implemented in POWHEG and MC@NLO, including W lν l decay. Distribution sample: used to estimate showering and hadronization uncertainties; bq bq W process being implemented.

18 t tb b at NLO: background for t th,h b b. [ ] dσ fb dm b b GeV. 5 [ ] dσ fb dp T,b GeV [Bredenstein, Denner, Dittmaier, Pozzorini, arxiv:95., arxiv:.46] pp t tb b + X NLO LO m b b > GeV m b b [GeV] pp t tb b + X NLO LO [Bevilacqua, Czakon, Papadopoulos, Pittau, Worek, arxiv:97.473] 35 4 dσ NLO dσ LO [ ] dσ fb dp T,b GeV pp t tb b + X m b b > GeV m b b [GeV] pp t tb b + X NLO LO 35 4 best central scale choice: µ = m t p b T p b T hard b jet often from initial state gluons. 5 m b b > GeV p T,b [GeV] m b b > GeV 5 5 p T,b [GeV] different from t th (Bredenstein, et al.)

19 Important to observe that: effect of jet veto on extra light jet: σ [fb] pp t tb b + X NLO LO dσ NLO dσ LO.5.5 pp t tb b + X m b b > GeV 5 5 p jet,veto [GeV].5 m b b > GeV 5 5 p jet,veto [GeV] regime of boosted Higgs: dσ d R [fb] b b pp t tb b + X.5 NLO LO m b b > GeV dσ d R [fb] b b pp t tb b + X.5 NLO LO p T,b b > GeV dσ NLO dσ LO pp t tb b + X p T,b b > GeV R b b R b b R b b

20 t tb b distributions: signal vs background, at NLO (Bevilacqua, Czakon, Garzelli, van Hameren, Papadopoulos, Pittau, Worek, arxiv:3.4, Les Houches 9) t tb b background: LO and NLO; t th with H b b: LO and NLO, calculated in NWA (valid for small M H ); to be revisited within the Higgs Cross Section WG (exclusive studies); signal now available in amc@nlo ( see also Maltoni s talk)

21 SM Branching Ratios Branching ratios - ττ bb gg WW ZZ tt LHC HIGGS XS WG uncertainties: - γγ cc Zγ theoretical (QCD, EW) parametric (m c, m b,...) linearly combined. -3 Tools: 3 5 M H [GeV] HDECAY [Djouadi, Kalinowski, Müllheitner, Spira] Prophecy4f [Bredenstein, Denner, Dittmaier, Mück, Weber] EW-NLO corrections to H γγ and H gg [Actis, Passarino, Sturm, Uccirati]

22 Strategy (from D. Rebuzzi s talk, BNL, May ) Calculate decay partial width as accurate as possible for each decay mode. Calculate branching ratio from full set of partial width. Define Higgs total width as Γ H = Γ HDECAY Γ HDECAY ZZ Γ HDECAY WW +Γ Profecy4f 4f where Γ Profecy4f 4f = Γ H WW 4f +Γ H ZZ 4f +Γ WW/ZZ int Results (preliminary, to be compared with [Baglio, Djouadi, arxiv:.53]) Process QCD EW Total H b b..% % (M H 35 GeV) 3 4% H c c..% % (M H 35 GeV) 3% H ττ % (M H 35 GeV) 3 6% H t t 5% 5% (M H 5 GeV) 5 % H WW/ZZ 4f < %.5% (M H 5 GeV) % H gg % % 5 7% H γγ <.5% < % %

23 Conclusions and Outlook We are living through a new era in Higgs-boson physics: looking for direct evidence. Higgs-boson precision physics has given a first coherent set of predictions for inclusive observables: Higgs-boson production cross sections and branching ratios. Short term: study exclusive observables, including decays, background processes, and experimental cuts. Long term: carry through a precision program that also include measurements of Higgs-boson properties, to identify possible candidates: the LHC will play an important role but need very high luminosity; LHC measurements will be important indications but are intrinsically model dependent; a high energy Linear Collider could be the best if not the only environment to complete and conclude the investigation of EWSB.