170 likes | 290 Views
Direct di- g Production @ Tevatron. Liang HAN. University of Science & Technology of China (USTC). On behalf of the & Collaborations. 1 st motivation: SM Higgs search@Tevatron. Production. 1. Gluon-gluon fusion gg H. Gluon-gluon fusion. W/Z associate.
E N D
Direct di-g Production @ Tevatron Liang HAN University of Science & Technology of China (USTC) On behalf of the & Collaborations
1st motivation: SM Higgs search@Tevatron • Production 1. Gluon-gluon fusion ggH Gluon-gluon fusion W/Z associate 2. W/Z associate qqW/Z+H W fusion (VBF) 3. W fusion (VBF) qqqq+H • Searching for light Higgs(~130GeV) • “inaccessible” ggHbb, S/B(QCD) ~ 10-9 135GeV • “common” qqbarW(ln)/Z(ll,nn) + H(bb) • “extra” ggHgg ,with Br(Hgg) ~ 0.2% Direct photon pair production at O(1)fb L.Han@Moriond-QCD 2010
SM prediction on direct diphoton production • Quark annihilation: LO (a2EM): NLO (asa2EM): virtual + real emission infra-safe + (ISR) (Box) • Gluon fusion: LO (a2sa2EM): gluon PDF density enhancement at low mass + • Fragmentation: FSR collinear singularity Suppressed by: • photon isolation • pT(gg)<M(gg) (no theory) L.Han@Moriond-QCD 2010
2nd motivation: precise test of QCD predictions • RESBOS, Phys. Rev. D 76, 013009 (2007) : + Quark Scattering qqbarggand Gluon Fusion ggggup to NLO + Fragmentation at LO, with additional NLO approximation + Resummation of soft/collinear terms of initial gluons up to all orders, cancelling divergence at NLO as pT(gg)0 • DIPHOX, Eur. Phys. J. C 16, 311 (2000) : + qqbargg up to NLO + ggggat LO + Fragmentation up to NLO + asymmetry di-photon pT(g1) > pT(g2) • PYTHIA, Comp. Phys. Comm. 135, 238(2001) : + qqbarggand ggggat LO + Resummation via parton shower L.Han@Moriond-QCD 2010
CDF results DIPHOX RESBOS • First di-g measurement @ Tevatron: 207pb-1, PRL 95, 022003 (2005) PYTHIA pT(g1)>14GeV, pT(g2)>13GeV; |h1,2|<0.9; ETiso<1GeV Mgg (GeV) pTgg (GeV) Dfgg (rad) Reasonable agreements between data and QCD predictions in different region : • Low pT (gg)~0GeV and Df~p, DIPHOX unstable due to the lack of resummation • Bump of pT(gg)~30GeV dominated by events of Df<p/2 andpT(gg)>M(gg), • described in DIPHOX as final state radiation + Fragmentation on the same quark L.Han@Moriond-QCD 2010
D0 di-photon measurement • D0 analysis based on 4.2fb-1 data: • pT(g1)>21GeV, pT(g2)>20GeV, |h1,2|<0.9, dR(1,2)>0.4 • Isolation requirement(jet and Fragmentation) + track veto(electron) • pT(gg)<M(gg) remove Fragmentation, reduce theoretical uncertainty • Neutral Network discriminator ONN to separate g from EM-like jet Loose L.Han@Moriond-QCD 2010
Background composition • Electron misidentified in Drell-Yan Z/g*ee : • Estimated with GEANT simulation, normalized up to NNLO and 4.2fb-1 • Jet-misidentified ing+jet and jet+jet : • Split data(Zee deducted) into 4 groups based on tighter ONN normalization + Npp : both pass + Npf : leading passes, trailing fails + Nfp : vice-versa + Nff : both fail Loose Tight 4×4 g/j ONN>0.6 efficiency matrix • Line shapes estimated by reversing ONN<0.1 L.Han@Moriond-QCD 2010
Differential cross section • Theoretical predictions: + RESBOS and DIPHOX, with CTEQ6.6M, mR=mF=mf=Mgg + PYTHIA 6.420 with CTEQ5L • Data.vs.MC comparison: Mgg (GeV) pTgg (GeV) Dfgg (rad) cosq*=tanh[(h1-h2)/2] • RESBOS with resummation demonstrates better agreement with data • data shows harder pT(gg) and excess in low Df (gg) L.Han@Moriond-QCD 2010
Double-differential cross section: • The pT(gg) inconsistence occurs in Mgg< 50GeV region, where the gluon fusion is significant. NNLO correction to gggg at low mass? • distributions tell the same story L.Han@Moriond-QCD 2010
Systematic uncertainty • Dominated by uncertainty of di-photon purity, ~10-15%, followed by luminosity ~ 6% • the accuracy is around O(1)fb, statistics are close to systematic L.Han@Moriond-QCD 2010
Impact on Higgs search • Reducible background (Zee, g+j, jj)subtracted, sideband fitting into signal region • Combine all signal channels (ggH, W/Z+H, VBF) to increase sensitivity D0: CDF: L.Han@Moriond-QCD 2010
Summary • Direct di-g production at Tevatron has been studied both by CDF and D0 • Data are compared with theoretical predictions, RESBOS, DIPHOX and PYTHIA. None of these calculations provides full description of data in all kinematic regions. • DIPHOX treats the Fragmentation better; impose pT(gg)<M(gg) would reduce the discrepancy to RESBOS; • RESBOS, with NLO gggg, gives the best agreement with data; hints the need of NNLO corrections for low mass region (<50GeV) • Provide extra the sensitivity to SM Higgs search in the most interested mass region ~130GeV L.Han@Moriond-QCD 2010
Backup Slides L.Han@Moriond-QCD 2010
CDF & DØ • Calorimeter : fine granularity and good energy resolution Isolation requirement to suppress fragmentation • CDF : Dh×Df ~ 0.1 ×0.26 • D0 : Dh×Df ~ 0.1 ×0.1 • Preshower: to distinguish g vs. neutral jets • D0 : lead + scintillating strip Central Preshower (CPS) • CDF : Preshower detector + shower maximal CES Shower shape difference between single g and multi-g from neutral hadron (e.g. p0gg) L.Han@Moriond-QCD 2010
+ Input : Preshower & Calorimeter shower shapes + tracker activities + Training : MC EM-like jet vs. g + Validation : data Zllg FSR ISR FSR Loose L.Han@Moriond-QCD 2010
Double-differential cross section: L.Han@Moriond-QCD 2010