1 / 31

Low-x QCD with forward jets in ATLAS, CMS & LHCb

EDS'09 International Conference CERN, 3 rd July 2009 David d'Enterria ICC-UB, Barcelona. Low-x QCD with forward jets in ATLAS, CMS & LHCb. Overview. 1. Introduction. Physics motivation: (1) Constrain low-x gluon PDF via fwd jets d  /dp T (3<| h |<5): data vs. pQCD

saman
Download Presentation

Low-x QCD with forward jets in ATLAS, CMS & LHCb

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. EDS'09 International Conference CERN, 3rd July 2009 David d'Enterria ICC-UB, Barcelona Low-x QCD with forward jetsin ATLAS, CMS & LHCb

  2. Overview • 1. Introduction. Physics motivation: • (1) Constrain low-x gluon PDF via fwd jets d/dpT (3<|h|<5): data vs. pQCD • (2) Study low-x QCDevolution via forward-backward “Muller-Navelet” dijets (cross-sections, azimuthal distributions, ...) • 2.Jet reconstruction in the LHC forward detectors: • - Acceptances (ATLAS, CMS, LHCb) • - Jet perfomances: pT, h, f resolutions & reco efficiencies • 3. Results (full-sim+reco, CMS): • 4. Summary (1) Inclusive forward jets dsjets /dpT vs. NLO: sensitivity to PDFs (2) “Muller-Navelet”dijets: dNMN-dijet/dpT vs. Dh, azimuthal decorrelation (dN/dDf, <cos(Df)>) vs. BFKL calculations.

  3. Parton densities at low-x • ■DIS e-p collisions probe distributions of partons in the proton: • ■Gluons only indirectly (F2“scaling violations”): Q 2= “resolving power” Bjorken x = momentum fraction carried by parton Q2 F2 ,F3 ,FL= proton structure functions, (y = inelasticity). J. Rojo et al. arXiv:0808.1231 xG(x,Q2) Large uncertainties ! for x<10-2 at moderate Q2 Q2 = 2 GeV2

  4. (x,Q2) evolution of PDFs • ■Q2 - DGLAP (kT-order'd emission): F2(Q2)~asln(Q2/Q02)n, Q02 ~1 GeV2[LT,coll.factoriz.] • ■x - BFKL (pL-ordered emission): F2(x) ~ asln(1/x)n[uPDFs, kT-factoriz.] • ■Linear equations­ single parton radiation/splitting ­ cannot work at low-x (i) Too high gluon density: nonlinear gluon- gluon fusion balances branchings. (ii) pQCD (collinear & kT) factorization assumptions invalid (HT, no incoherent parton scatt.). (iii) Violation of unitarity even for Q2>>L2 (too large perturbative cross-sections) ■Non-linear effects may “leak” to higher Q2 (“geometric scaling”) region

  5. Experimental access to low-x gluon PDF • ■Perturbative processes: • ‣e-p:∂lnF2/∂lnQ2, FL,heavy-Q, • diffractive QQ: • ‣p-p: (di)jets, prompt g, heavy-Q, ...: _ ... (di)jets (y=4) DdE, arXiv:0708.0551 ‣Forward production: x2s/2 x1s/2 x2min ~ pT/√s ·e-y= xT·e-y Every 2-units of y, xmindecreases by ~10

  6. Low-x studies at the LHC • ■p-p @ 14 TeV : • (1) At y=0, x=2pT/√s~10-3 (domain probed at HERA,Tevatron). Go fwd. for x<10-4 • (2) Saturation momentum: Qs2~ 1 GeV2 (y=0), 3 GeV2 (y=5) • (3) Very large perturbative cross-sections: Prompt g Drell-Yan Jets Heavy flavour _ ep, pp W,Z production ? LHC forward rapidities: e.g. y ~ 6, Q ~ 10 GeV x down to 10-6 !

  7. Forward detectors at the LHC ■Most of phase-space y~2✕ln(√s)/mp~20 covered: 1st time in a collider ! ■Calorimeters: CMS/ATLAS: up to |h|~6.6 LHCb: 2<h<5 ■Muon spectrometers: LHCb: 2<h<5 ALICE: 2.5<h<4 ■Trackers/PID/2nd vtx: LHCb: 2<h<5 (TOTEM partially) DdE, arXiv:0806.0883 x2min ~ pT/√s ·e-y= xT·e-y Every 2-units of y, xmindecreases by ~10

  8. x2 ~ 10-4 log10(x1,2)‏ Observable (1): Gluon PDF via fwd. jets • ■Jets with pT ~20-100 GeV/c at forward rapidities (3<||<5) probe x2 ~10-4: PDF cross-checks and/or global-fit analyses. varying PDFs: ~60% yield diffs. at pT~ 40GeV/c S.Cerci,DdE: arXiv:0812.2665

  9. Observable (1): Gluon PDF via fwd. jets ■Jets with pT ~20-100 GeV/c at forward rapidities (3<||<5) probe x2 ~10-4: PDF cross-checks and/or global-fit analyses. varying PDFs: x2 ~ 10-4 ~60% yield diffs. at pT~ 40GeV/c log10(x1,2)‏ S.Cerci,DdE: arXiv:0812.2665

  10. Observable (2): Low-x QCD via fwd-back. dijets ■Mueller-Navelet dijets with large y separation very sensitive to low-x QCD evolution (testing ground for BFKL): ■Increased azimuthal decorrelation with increasing Dy (w.r.t. DGLAP collinear-factorization): jet1 y~10 Extra BFKL radiation between the 2 jets smooths out back-to-back topology jet2 A.H.Mueller, H.Navelet, NPB282 (1987)727 BFKL (LHC) ~80% decorrelation [DelDuca, Schmidt], [Orr, Stirling] [A.Sabio-Vera, F.Schwennsen] [C.Marquet, Royon] [E. Iancu et al.] Dh

  11. Observable (2): Low-x QCD via fwd-back. dijets ■Mueller-Navelet dijets with large y separation very sensitive to low-x QCD evolution (testing ground for BFKL): ■Increased azimuthal decorrelation with increasing Dy (w.r.t. DGLAP collinear-factorization): jet1 y~10 Extra BFKL radiation between the 2 jets smooths out back-to-back topology jet2 A.H.Mueller, H.Navelet, NPB282 (1987)727 _ Tevatron, pp √s = 1.8 TeV NLO BFKL [DelDuca, Schmidt], [Orr, Stirling] [A.Sabio-Vera, F.Schwennsen] [C.Marquet, Royon] [E. Iancu et al.] [D0 Collab, PRL77(97)595]

  12. 2. Forward jet reconstructionperfomances at the LHC

  13. LHC forward (EM/HAD) calorimeters 3 < |h|< 5 2 < h< 5 3 < |h|< 6.6 Note: also tracking !

  14. ATLAS/CMS/LHCb: Jet algorithms 3 std. jet algos (ICone, kT,SISCone) available in official codes: LHCb: Calo+Tracks. Secondary vtx. required (focus on b-jets so far).  ATLAS/CMS: CaloJets only (fwd). Std jet energy corrections applied: (*) (*) Note: smallish radii chosen (R=0.5~D=0.4) to minimize UE (fwd.) CorrCaloJet

  15. ~20% (pT~20 GeV/c), ~12% (pT>100 GeV/c) ATLAS/CMS: forward jet resolutions [CMS plots. Similar results for ATLAS] Position h-f resolution: pTresolution : ~0.04 (pT~20 GeV/c) ~0.025 (pT>100 GeV/c)  ICone~Fast-kT~SISCone.

  16. ATLAS/CMS: fwd. jet reco efficiency Good reconstruction efficiency above pT ~35 GeV/c: S.Cerci,DdE: arXiv:0812.2665 ATLAS-TDR CERN-OPEN'08-20 ~95% ~80% CMS preliminary ~10% fake rate pT~20 GeV/c: ~50% fake jets Fake jets <35 GeV/c: clustered underlying event, beam-remnants, noise ...

  17. LHCb: fwd. jet reco performances Marco Musy [LHCb] Focus onb-jets (H  bbar in VH associated production) so far. Algos: kT (D=0.75) & seed cone (R=0.7). Neural-network for b-jet selection Calo saturation resolution ~14 mrad  Ejet calbrated (GeV) energy correction Theta of true vertices Ejet true (GeV) Theta of reco vertices Nice jet-reco capabilities (tracking): vertexing, position resol., ... Calo satu- ration is an issue only at high-pT. Interesting potential for fwd light-q,g jets.

  18. 3. Fwd. jet spectrum (CMS):PDF sensitivity

  19. Fwd jet spectrum (1 pb-1): algos comparison S.Cerci,DdE: arXiv:0812.2665 Invariant cross-sections : ICone, SiSCone, Fast-kT Yields: ~1M fwd. jets with pT>35 GeV/c CMS preliminary CMS preliminary SISCone & IterativeCone yields are very similar. Fast-kT is 20%-25% lower than cone algorithms below pT80 GeV/c

  20. 10% 10%+5% Fwd. jet spectrum (1 pb-1): uncertainties S.Cerci,DdE: arXiv:0812.2665 Invariant cross-sections: CorrCaloJets vs. NLO-MRST03,CTEQ6M  Small point-to-point errors: statistical + 2%-5% E-resolution unsmearing Main systematic uncertainty: jet-energy-scale (JES). 2 calibration scenarios: FastNLO: K.Rabbertz et al. hep-ph/0609285 CMS preliminary

  21. Fwd. jet spectrum (1 pb-1): PDF sensitivity S.Cerci,DdE: arXiv:0812.2665 Fractional x-section difference: CorrCaloJets vs. NLO-MRST03,CTEQ6M  Small p2p errors (stat.+E-resol) butlarge pT-corr (JES) errors: ~40-50%  JES scenario-1 (>50% errors): No PDF sensitivity  JES scenario-2 (30%-40% errors): PDF sensitivity for <60 GeV/c

  22. 4. ”Muller-Navelet” dijets (CMS): yields, azimuthal decorrelation

  23. Muller-Navelet dijets: kinematic cuts Jet1 y~10 MN-like dijet event Jet2 2 forward jets : 3. < |h| < 5. Moderately hard: pT > 35 GeV (good parton-jet match, trigger effic.) Similar pT (minimize DGLAP rad.): |pT1-pT2| < 5 GeV Jets in opposite:h1*h2 < 0. HF HF Underlying event

  24. MN dijets: Expected yields (1 pb-1) Stats: ~5000(200) Mueller-Navelet-type dijets separated by Dh~6 (9). pT 1,2> 35 GeV/c Dh~7.5 Dh~6.5 S.Cerci,DdE: arXiv:0812.2665 enough stats for detailed studies of Dy-evolution in |Dh|=6-9: yields, Df decorrelation Dh~8.5 Dh~9.5

  25. MN dijets: azimuthal decorrelation S.Cerci,DdE: arXiv:0812.2665 BFKL predictions (parton-level): PYTHIA/HERWIG vs BFKL: 〈cos (p - Df)〉 [A.Sabio,F.Schwennsen] [C.Marquet, Royon] HFs range p - Df HERWIG more decorrelation (~15%) than PYTHIA but ~20% less than BFKL analytical estimates (parton showering & hadroniz. will still increase this effect).  Extra radiation enhances azimuthal (back-to-back) decorrelation.

  26. Summary 1. Physics motivation: Forward (di)jets in p-p at 14 TeV sensitive to: (i) low-x gluon PDFs, (ii) non-DGLAP (BFKL, saturation) QCD evolution. 2. Jet recoperformances in forward calorimeters: ✶Similar for all algos (cone,kT) for smallish jet radius (R=0.5~D=0.4) ✶Good reconstruction for pT >35 GeV/c. 3. Forward jets single spectrum: ✶ Large stats. (~1M jets, 1pb-1) but large syst. error (>30%) from JES. ✶Sensitivity to PDFs differences (pT~35-60 GeV/c) iff JES controlled below 5%. 4. “Muller-Navelet” dijets: ✶ Stats (ℒ~1 pb-1): ~5000 (200) dijets separated by Dh~6 (9). ✶ “Normal” azimuthaldecorrelations: ~10-20% (hadronization+FS radiation) ~10-20% (exp. reco effects). HERWIG ~15% more decorrelation than PYTHIA. ✶Enhanced BFKL decorrelation should be identifiable in the data (<cos(Df)>)

  27. Backup slides

  28. Gluon saturation hints at HERA • ■DGLAP fits most of e-p data. Saturation models explain better a few cases: (2) flat sdiffract/stot vs energy (3) Long. struc. function (1) “Geometric scaling” sg*p Q0 =1GeV λ ~ 0.3 e.g. R.S.Thorne arXiv:0808.1845 Golec-Biernat-Wusthoff PRD60 114023 (1999) t = Q2/Q2s Gluon (FL) at NLO becomes negative for Q2~2 GeV2 at low-x Diffract. & total x-sections similar W dependence  pQCD: stot~W2l sdiff~W4l Inclusive DIS x-section depends on single scale Q2/Qs2 for x < 0.01

  29. CMS Physics TDR-I CMS Physics TDR-I CMS-PAS JME-07-003 ~0.05 (pT~20 GeV) ~0.020 (pT>100 GeV) ~0.050 rads (pT~20 GeV), ~0.020 rads (pT>100 GeV) HF(3<||<5) position (h,f) resolution Good hresolution Good fresolution: (Note: phi resolution better at fwd than central rapidities) (Note: eta resolution better at fwd than central rapidities) Note: resolution is better than at fwd. rapidities (boosted jet) !

  30. MN dijets: 〈cos n(p - Df)〉 Average decorrelation vs. Dh: Parton-level CorrCaloJet GenJet LO back-to-back topology Point-to-point errors: Statistical (dominant) Dijet azimuthal decorrelation (<cos(p - Df)>) increased by: ~10-20% due parton hadronization+FS radiation effects. ~10-20% due experimental reconstruction effects.

  31. MN dijets: 〈cos(nDf)〉 at Leading-Log Average cos(nDf) vs. Dh: PYTHIA vs BFKL BFKL (LL) calculations CorrCaloJet (PYTHIA) [L.H. Orr W.J. Stirling PLB436(1998)37] Increasing (up to ~80%) azimuthal decorrelation with jet rapidity separation Small (~20%) azimuthal decorrelation

More Related