Electroweak corrections with

Transcription

1 Electroweak corrections with Marco Zaro, LPTHE-UPMC in collaboration with S.Frixione, V.Hirschi, D.Pagani, H.-S. Shao

2 Motivation No clear sign of new physics seen at LHC Run I, good agreement between data and SM; Run II has just started gluino mass [GeV] New physics state will likely manifest as tiny deviations from the SM: Accurate predictions strongly needed LSP mass [GeV] ~ g- ~ g production, CMS Preliminary s = 8 TeV ICHEP 214 Observed SUSY Observed -1 σ theory Expected m(gluino) - m(lsp) = 2 m(top) ~ g t t χ 1-1 SUS lep (E T +H T ) 19.5 fb -1 SUS lep (razor) 19.3 fb SUS lep (n jets -1 6) 19.3 fb SUS lep (OS+b) 19.7 fb -1 SUS lep (SS+b) 19.5 fb -1 SUS lep (3l+b) 19.5 fb -1 2

3 Accurate predictions in a nutshell Expand the cross-section as a series in the couplings d LO NLO NNLO = d apple 1+ s s QCD EW Strong coupling dominates, but non-qcd effects must be accounted to achieve precision Roughly speaking: NLO EW ~ NNLO QCD 3

4 EW corrections: how to 4

5 EW corrections: how to 2 2 LO In the general case, several coupling combinations contribute to a given process at LO 4

6 EW corrections: how to 2 2 LO In the general case, several coupling combinations contribute to a given process at LO Typically the LO is identified with the blob featuring the largest power of αs 4

7 EW corrections: how to 2 2 LO 2 NLO In the general case, several coupling combinations contribute to a given process at LO Typically the LO is identified with the blob featuring the largest power of αs NLO QCD corrections can be computed by attaching QCD particles to the LO 4

8 EW corrections: how to 2 2 LO 2 2 NLO In the general case, several coupling combinations contribute to a given process at LO Typically the LO is identified with the blob featuring the largest power of αs NLO QCD corrections can be computed by attaching QCD particles to the LO NLO EW corrections can be computed by attaching EW particles to the LO 4

9 EW corrections: how to 2 2 LO 2 2 NLO In the general case, several coupling combinations contribute to a given process at LO Typically the LO is identified with the blob featuring the largest power of αs NLO QCD corrections can be computed by attaching QCD particles to the LO NLO EW corrections can be computed by attaching EW particles to the LO and attaching QCD particles to the LO with one less power of αs 4

10 In practice: EW corrections are a more complex problem, yet similar to QCD ones Lot of book-keeping (in particular on coupling orders) Loops: Renormalize the full model Higher rank loop integrals (still R D) Interfere QCD loops with EW borns (and viceversa) Real emissions: keep track of the splitting type to generate the correct born-like counterterms Modern techniques for QCD corrections can be extended to tackle EW ones 5

11 EW corrections in Loop computation: MadLoop Hirschi et al, arxiv: Uses OPP or TIR UV renormalisation in the α(mz) or Gμ scheme (model feature) Well advanced validation for complex-mass scheme IR subtraction and integration: MadFKS Frederix et al, arxiv: Code modified to account for more than one type of counterterms All kind of EW/QCD splittings are accounted for New version will be able to keep track of contributions from different coupling order combinations, and exact scale uncertainties via reweight Matching with parton shower Work in progress 6

12 Physics results Appetizer: t t production at the LHC Main course: t t + W/Z/H production at the LHC CMS Projection Expected uncertainties on Higgs boson couplings -1 3 fb at -1 3 fb at s = 14 TeV Scenario 1 s = 14 TeV Scenario 2 κ γ κ W κ Z κ g κ b κ t κ τ expected uncertainty 7

13 EW corrections to t t production Motivation ATLAS and CMS see some anomaly on the top pt distribution and t t invariant mass Top Top pm T : a very interesting variable tt pt Data are softer than NLO QCD MonteCarlos (up to 3-4%) Heavy resonances decaying to tt, see Sabine s talk yesterday CERN-THESIS \ CERN-THESIS \ tt dominant background Is it an EW effect? The other side of the coin Both ATLAS and CMS see lower diffxsec Heavy resonances hints might be hidden in than predicted at high pt,top mtt differential measurements deviations Accurate ATLASmodelling of SM mtt shape Powheg+Pythia is8 consistent between improves dominant background description & CMS

14 Contributions to the cross-section αs 2 αsα α 2 LO LO QCD LO EW NLO QCD NLO EW NLO αs 3 αsα 2 αs 2 α α 3 LO2 has only gγ and bb initial states; dominant γ-initiated contribution, need for PDFs with photons NLO2 formally also includes heavy boson radiation (HBR). HBR not included for t t 9

15 EW corrections to t t production 1 2 tt - production at the 7 TeV LHC LO 1 +NLO 1 (QCD) 1 2 tt - production at the 7 TeV LHC LO 1 +NLO 1 +LO 2 σ per bin [pb] LO QCD QED NLO LO 1 +NLO 1 +LO 2 +LO 3 LO 1 +NLO 1 (QCD) LO 1 +NLO 1 +LO 1 +NLO 2 +NLO 1 +LO 2 LO 1 +NLO 1 +LO 2 +LO 3 LO 1 +NLO 1 +LO 2 +NLO 2 +LO 3 +NLO 3 +NLO 4 LO 1 +NLO 1 +LO 2 +NLO 2 LO 1 +NLO 1 +LO 2 +NLO 2 +LO 3 +NLO 3 +NLO 4 MadGraph5_aMC@NLO σ per bin [pb] LO 1 +NLO 1 (QCD) LO 1 +NLO 1 +LO 2 LO 1 +NLO 1 +LO 2 +LO 3 LO 1 +NLO 1 +LO 2 +NLO 2 LO 1 +NLO 1 +LO 2 +NLO 2 +LO 3 +NLO 3 +NLO 4 MadGraph5_aMC@NLO 1.2 ratio over LO+NLO QCD and QCD scale unc LO+NLO QCD + LO EW no γ.8 LO+NLO QCD + LO+NLO EW no γ Zoom ratio over LO+NLO QCD and PDF unc. 1.2 ratio over LO+NLO QCD and QCD scale unc LO+NLO QCD + LO EW no γ.8 LO+NLO QCD + LO+NLO EW no γ Zoom ratio over LO+NLO QCD and PDF unc p T (t) [GeV] M(tt - ) [GeV]

16 EW corrections to t t production Comments EW corrections account at most -1% at large pt, -5% at large mass Photon effect as large as EW corrections, but almost 1% uncertain Subleading corrections (LO3, NLO3,4) very small 11

17 EW corrections to t t + W/Z/H Motivation t t H offers unique access to top Yukawa coupling Unlike QCD, EW effects introduce non ~yt 2 dependence of the cross-section Expected accuracy on yt at Run II: 1-5% with 3-3 fb -1 Searches in the boosted scenario: EW corrections enhanced by Sudakov logs (log(pt/mw)) CMS Projection Expected uncertainties on Higgs boson couplings κ γ 3 fb -1 at -1 3 fb at s = 14 TeV Scenario 1 s = 14 TeV Scenario Plehn, Salam, Spannowsky, arxiv: / tot d /dp T κ W κ Z κ g 1-2 tth: p T,t κ b κ t κ τ 1-3 tth: p T,H expected uncertainty Wjj: p T,j WH: p T,H p T [GeV]

18 EW corrections to t t + W/Z/H Motivation t t V measured at Run I, still large statistical uncertainties EW corrections needed for high-precision Run II measurements Multilepton/leptons+jets signatures: t t V background to many BSM searches and to t t H 13

19 Setup more in Frixione, Hirschi, Shao, Zaro, arxiv: EW corrections computed in the α(mz) scheme (G μ also available) Particle masses: mt=173.3 GeV mw=8.385 GeV mh=125 GeV mz= GeV NNPDF2.3 QED PDF, quoted Ren./Fac. scale choice: LO+NLO QCD scale uncertainties in the range µ = H T µ µ R,µ F 2µ Two scenarios: inclusive and boosted: pt(t, t,w/z/h)>2 GeV 14

20 Results for t t H and t t Z: total rates (within boosted cuts) t th : σ(pb) 13TeV t tz : σ(pb) 13TeV LO QCD NLO QCD LO EW LO EW no γ NLO EW NLO EW no γ HBR ( ) ( ) ( ) ( ) ( ) ( ) ( ) LO QCD NLO QCD LO EW LO EW no γ NLO EW NLO EW no γ HBR ( ) ( ) ( ) ( ) ( ) ( ) ( ) t th : δ(%) NLO QCD LO EW LO EW no γ NLO EW NLO EW no γ TeV ± 2.8 ( ± 4.5) 1.2 ±.9 (2.8 ± 2.).4 ±. (.2 ±.) 1.2 ±.1 ( 8.2 ±.3) 1.4 ±. ( 8.5 ±.2) t tz : δ(%) NLO QCD LO EW LO EW no γ NLO EW NLO EW no γ TeV ± 2.9 ( ± 4.7). ±.7 (2.1 ± 1.6) 1.1 ±. (.3 ±.) 3.8 ±.2 ( 11.1 ±.5) 4.1 ±.1 ( 11.5 ±.3) HBR.89 (1.87) NLO EW correction have modest impact on inclusive xsect, but can be important in the boosted regime (same order of QCD uncertainties) Boosted regime enhances photon contribution in LO-EW HBR.96 (2.13) HBR contributions remain small 15

21 Results for t t H and t t Z: distributions 1-1 tt - Z production at the 13 TeV LHC 1-2 tt - H production at the 13 TeV LHC boosted cuts: p T (t), p T (t - ), p T (H) > 2 GeV LO QCD LO QCD per bin [pb] ratio over LO QCD; scale unc ratio over LO QCD; PDF unc relative contributions NLO QCD LO+NLO EW LO+NLO QCD LO+NLO QCD+EW LO+NLO QCD+EW, no LO+NLO EW, no HBR p T (Z) [GeV] MadGraph5_aMC@NLO σ per bin [pb] ratio over LO QCD; scale unc ratio over LO QCD; PDF unc relative contributions NLO QCD LO+NLO EW LO+NLO QCD LO+NLO QCD+EW LO+NLO QCD+EW, no γ LO+NLO EW, no γ HBR p T (H) [GeV] MadGraph5_aMC@NLO 16

22 Results for t t W: total rates (within boosted cuts) t tw + : σ(pb) LO QCD NLO QCD LO EW LO EW no γ 13TeV ( ) ( ) t tw : σ(pb) LO QCD NLO QCD LO EW LO EW no γ 13TeV ( ) ( ) NLO EW NLO EW no γ HBR ( ) ( ) ( ) NLO EW NLO EW no γ HBR ( ) ( ) st in table 1 3, 3 for (4.781 ttw production 1 4 ) t tw + : δ(%) NLO QCD LO EW LO EW no γ NLO EW NLO EW no γ HBR TeV ± 2.4 ( ± 3.1) 7.7 ±.2 ( 19.2 ±.7) 8. ±.2 ( 2. ±.5) 3.88 (7.41) t tw : δ(%) NLO QCD LO EW LO EW no γ NLO EW NLO EW no γ HBR TeV ± 2.8 ( ± 3.9) 6.7 ±.2 ( 18.3 ±.8) 7. ±.2 ( 19.1 ±.6) 6.5 (15.1) EW corrections larger than t t H/Z, in particular with boosted cuts HBR enhanced by parton luminosities: t t WW has gg, t t W only qq 17

23 Results for t t W: distributions tt - W - production at the 13 TeV LHC 1-1 tt - W + production at the 13 TeV LHC 1-2 LO QCD LO+NLO QCD LO QCD LO+NLO QCD per bin [pb] LO+NLO QCD+EW LO+NLO QCD+EW, no MadGraph5_aMC@NLO σ per bin [pb] LO+NLO QCD+EW LO+NLO QCD+EW, no γ MadGraph5_aMC@NLO ratio over LO QCD; scale unc. ratio over LO QCD; PDF unc. relative contributions NLO QCD LO+NLO EW p T (W - ) [GeV] LO+NLO EW, no HBR ratio over LO QCD; scale unc ratio over LO QCD; PDF unc relative contributions NLO QCD LO+NLO EW LO+NLO EW, no γ HBR p T (t) [GeV] 18

24 Conclusions NLO EW predictions will be very important for accurate physics at the LHC RunII Automation of EW corrections in well advanced Results (obtained automatically) for t t and t t X (X=H/Z/W) EW corrections seem not to explain the t t anomaly seen by ATLAS and CMS t t X (in particular t t W) can receive large corrections, specially in boosted regimes 19

25 Thank you 2