Drell-Yan processes at the LHC

Transcription

1 Drell-Yan processes at the LHC Alessandro Vicini University of Milano, INFN Milano Firenze, October 1st 27 with: C. M. Carloni Calame, G. Balossini, G. Montagna, O. Nicrosini, F. Piccinini, M. Moretti, M. Treccani 1

2 Outline relevance of Drell-Yan processes and motivation for precision studies Charged Current and Neutral Current processes EW O(α) corrections matched with higher order QED corrections photon induced processes impact of the EW corrections on several observables preliminary study to extract MW from the ratio of W and Z distributions combining QCD and EW radiative corrections 2

3 The Drell-Yan processes p X easy detection high pt lepton pair or high pt lepton + missing pt typical cuts at the LHC (central detector region) u d! W +! l l + p,l and p,ν > 25GeV, η l < 2.5 p X large cross section at LHC σ(w) = 3 nb i.e. 3 1^8 events with L=1 fb^-1 at LHC σ(z) = 3.5 nb i.e ^7 events with L=1 fb^-1 no statistical limitation to perform high precision EW measurements lepton distributions W mass and width W transverse mass ratios W/Z distributions pdf validation collider luminosity detector calibration total cross section W, Z rapidity lepton pseudo-rapidity acceptances W, Z mass distributions! (nb) proton - (anti)proton cross sections! tot! b! jet (E T jet > "s/2)! W! Z! jet (E T jet > 1 GeV)! t! jet (E T jet > "s/4)! Higgs (M H = 15 GeV)! Higgs (M H = 5 GeV) Tevatron "s (TeV) LHC events/sec for L = 1 33 cm -2 s -1 3

4 Relevance of a precise W mass measurement Sensitivity to the precise value of the Higgs boson mass or to SUSY particles 4

5 Relevance of a precise W mass measurement Sensitivity to the precise value of the Higgs boson mass or to SUSY particles 1 3 New world average: m H!GeV" m!gev" m W (prel.)!gev" C. Hays, University of Oxford 4

6 Relevance of a precise W mass measurement Sensitivity to the precise value of the Higgs boson mass or to SUSY particles 1 3 New world average: m H!GeV" m!gev" m W (prel.)!gev" W-Boson Mass [GeV] C. Hays, University of Oxford Top-Quark Mass [GeV] TEVATRON ±.39 LEP ±.33 Average ± m W [GeV] χ 2 /DoF: 1.1 / 1 NuTeV ±.84 LEP1/SLD ±.32 LEP1/SLD/m t 8.36 ±.2 CDF 17.1 ± 2.2 D 172. ± 2.4 Average 17.9 ± 1.8 χ 2 /DoF: 9.2 / 1 LEP1/SLD LEP1/SLD/m W /Γ W m t [GeV] To ensure that top and W mass measurements have the same weight in the SM EW fit, the experimental errors should be related as: m W m t 4

7 Precision measurement of EW observables W-Boson Mass [GeV] TEVATRON ±.39 LEP ±.33 Average ±.25 χ 2 /DoF: 1.1 / 1 NuTeV ±.84 LEP1/SLD ±.32 LEP1/SLD/m t 8.36 ± m W [GeV] measurement M W error target at the LHC 15 MeV New projection with 1.5 fb -1 of data:!m W < 25 MeV with CDF C. Hays, University of Oxford Γ W measurement error at Tevatron: 87 MeV error target at the LHC: 3 MeV sin 2 θ lep eff measurement world average:.23122±.15 error target at the LHC:.14 5

8 Constraining the pdfs H.Stenzel - HERA-LHC workshop!"br[fb] x W + W - x 1 1!"BR[fb] 8 6 LHC pp # W(#e+$ + jet) NLO (MCFM) W + W -!"BR[fb] x LHC pp # W(#e+$ + jet) NLO (MCFM) W + W - 2 LHC pp # W(#e+$) NLO (MCFM) CTEQ61 ± uncertainty 4 CTEQ61 ± uncertainty MRST21 ± uncertainty 4 1 MRST21 ± uncertainty 2 2 CTEQ61 ± uncertainty Luminosity monitoring (?) Ratio W + /W Ldt = CTEQ ± uncertainty double ratio CTEQ/MRST MRST ± uncertainty % lepton 1 BR(W lν) 1 σ th (W ) N obs A W A W (ηl max ) = 1 σ ( tot) η max l dη l dσ (cuts) dη l Table 5: Total cross-sections and systematic uncertainties within the experimental η max acceptance for W/Z + jet processes. l the recent set CTEQ6.5 throws some shadows on this approach because of the CT large EQ (8%) induced theoretical uncertainties Ratio W + /W CTEQ ± uncertainty double ratio CTEQ/MRST MRST ± uncertainty % lepton Ratio W + /W - MRST21 ± uncertainty CTEQ ± uncertainty MRST ± uncertainty Fig. 18: Left: pseudo-rapidity distribution of the decay lepton from inclusive W +jet production and right: pseudo-rapidity of the associated leading jet. The bands represent the PDF-uncertainty. W + + jet W + jet Z + jet CTEQ61 [pb] P DF [pb] ±44.1 ±34.3 ±9.1 double ratio CTEQ/MRST % jet <=>?7,@&9#+%,#A?%#'8B,7%#'&%'&./12365 CD&E:FG CT EQ P DF [%] ±4.2 ±4.4 ±4.3 MRST [pb] MRST P DF [pb] ±17.6 ±14.8 ±3.67 A. Cooper-Sarkar - DIS27 MRST P DF [%] ±1.7 ±1.9 ±1.8 5H!/IJ pert [%] #8&9:;#7%#'8&$,'-&./1234./12345:&./12;44< 29 very promising in the long term (it requires a careful pdfs validation) 6

9 The quest for precision (I) transverse mass M W = i.e. how do we measure MW? 2p l pν (1 cos φ lν) reconstructed in the transverse plane jacobian peak at the W mass rather insensitive to QCD initial state radiation (e.g. ptw modeling) dn/dm T dn/dp T (e) no det., no p W with p W with det. and p W m T (GeV) p T (e) (GeV) Detector response effects strongly affect the distributions QED Final state radiation distorts the lepton p and transverse mass distributions affects the determination of M W muons O(α)corrections: M W = 168 ± 2 MeV electrons M W = 65 ± 2 MeV 7

10 The quest for precision (II) C.M. Carloni Calame et al., Phys. Rev. D69 (24) 3731 What is the effect of QED higher orders on the MW extraction? Shift induced in the extraction of MW from higher order QED effects (very simplified detector for muons and electrons) #$ W! µ " W! e " order % #$ W! µ " W! e " exponentiation M α W = 11 MeV M exp W = 1 MeV #M W (MeV) #M W (MeV) M W systematics In agreement with CDF estimates S. Malik@DIS27 8

11 QCD approximations and tools NLO/NNLO corrections to W/Z total production rate G. Altarelli, R.K.Ellis, M. Greco, G. Martinelli, Nucl.Phys.. B246 (1984) 12 R. Hamberg, W. L. van Neerven, T. Matsuura, Nucl.Phys. B359 (1991) 343 W. L. van Neerven and E.B. Zijstra, Nucl.Phys. B382 (1992) 11 Fully differential NLO corrections to l l (MCFM) J. M. Campbell and R.K. Ellis, Phys.Rev.D65:1137 Fully differential NNLO corrections to l l (FEWZ) C. Anastasiou et al., Phys.Rev. D69 (24) 948 K. Melnikov and F. Petriello, hep-ph/63182 resummation of LL/NLL p W /M W logs (RESBOS) C.Balazs and C.P. Yuan, Phys.Rev. D56 (1997) 5558 NLO ME merged with HERWIG PS (MC@NLO) S. Frixione and B.R.Webber., JHEP 26, 29 (22) LO Matrix Elements Monte Carlos (ALPGEN, SHERPA,...) matched with PS M.L.Mangano et al., JHEP 37, 1 (23) F. Krauss et al., JHEP 57, 18 (25) 9

12 EW results and tools W production O(α 2 S) O(α em ) Need to worry about EW corrections Pole approximation D.Wackeroth and W. Hollik, PRD 55 (1997) 6788 U.Baur et al., PRD 59 (1999) 132 Exact O(alpha) V.A. Zykunov et al., EPJC 3 (21) 9 S. Dittmaier and M. Krämer, PRD 65 (22) 737 DK U. Baur and D. Wackeroth, PRD 7 (24) 7315 WGRAD2 A. Arbuzov et al., EPJC 46 (26) 47 SANC C.M.Carloni Calame et al., JHEP 612:16 (26) HORACE Photon-induced processes S. Dittmaier and M. Krämer, Physics at TeV colliders 25 A. B.Arbuzov and R.R.Sadykov, arxiv: Multiple-photon radiation Z production only QED C.M.Carloni Calame et al.,prd 69 (24) 3731, JHEP 612:16 (26) HORACE S.Jadach and W.Placzek, EPJC 29 (23) 325 WINHAC U.Baur et al., PRD 57 (1998) 199 Exact O(alpha) U.Baur et al., PRD 65 (22) 337 V.A. Zykunov et al., PRD75 (27) 7319 C.M.Carloni Calame et al., to appear ZGRAD2 HORACE Multiple-photon radiation C.M.Carloni Calame et al., JHEP 55:19 (25) HORACE 1

13 The HORACE event generator developed by: C.M.Carloni Calame, G.Montagna, O.Nicrosini, A.Vicini exact O(α) radiative corrections matched with multiple photon radiation via QED Parton Shower true, fully exclusive event generator events saved in a Les Houches compliant form interfaced to LHAPDF package easy to interface to QCD showering programs like HERWIG or PYTHIA 11

14 Basics of the HORACE code (both CC and NC channels) LO calculation + QED LL corrections to all orders via Parton-Shower exact O(alpha) EW radiative corrections matching of Parton-Shower and exact results (no double countings) use MRST24QED: consistent description of initial state QED radiation photon density in the proton photon induced processes subtraction procedure of IS collinear divergences to all orders (α, m W, m Z ) The input parameters scheme, i.e. renormalization, gauge boson masses as input parameters which can be fitted from the data numerical evalution in the scheme (CC) or ( ) scheme (NC) G µ G µ + α(q 2 ) 12

15 Partonic processes at O(α) u d l + ν l (1γ) virtual corrections checks: UV finiteness IR finiteness (when combined with soft photon emission) use of two different gauges (Feynman and background) fixed W decay width necessary to describe the resonance region included in all tree-level propagators and at 1-loop in all the resonant logs: on-shell renormalization scheme large negative EW Sudakov logs real bremsstrahlung corrections checks: independence of total cross section of soft/hard separator e.m. gauge invariance initial state collinear logs regulated by quark masses large ISR corrections: radiative return to the W resonance photon-induced process: additional IS collinear divergences to be subtracted q q l + l (1γ) log(s m 2 W ) log(s m 2 W + iγ W m W ) similar computational path as in the CC case delicate treatment of the Z decay width in the virtual corrections photon-induced process: further additional IS collinear divergences to be subtracted 13

16 exact Matching exact O(α) parton-shower (PS) O(α) resummed PS O(α) partonic cross-section dσ α,ex dσ α,ex O and parton-shower results SV + dσ α,ex H dσ α,p S = [ Π S (Q 2 ) ] O(α) dσ + α 2π P (x)i(x)dx dc dσ dσ α,p S SV + dσ α,p S H dσ P S = Π S (Q 2 ) n= dˆσ 1 n! n i= ( α 2π P (x i) I(k i ) dx i d cos θ i ) Π S (Q 2 ) I(k i ) = (k i ) 2 exp ( α2π ( ) Q 2 1 ε log N j,l=1 η j η l m 2 p j p l (p j k i )(p l k i ) ) dx P (x) Sudakov form factor photon angular spectrum 14

17 exact Matching exact O(α) parton-shower (PS) O(α) O(α) partonic cross-section dσ α,ex dσ α,ex O and parton-shower results SV + dσ α,ex H dσ α,p S = [ Π S (Q 2 ) ] O(α) dσ + α 2π P (x)i(x)dx dc dσ dσ α,p S SV + dσ α,p S H resummed PS + exact O(α) Π S (Q 2 )F SV dσ matched = n= dˆσ 1 n! n i= ( α 2π P (x i) I(k i ) dx i d cos θ i F H,i ) F SV = 1 + dσα,ex SV dσα,p S SV F H,i = 1 + dσα,ex H,i dσ dσ α,p S H,i dσ α,p S H,i at O(α) it coincides with the exact calculation QED higher orders coincide with pure PS 15

18 The hadronic process at O(α) σ(pp l lx) = a,b 1 pp(p p) l lx dx 1 dx 2 q a (x 1 )q b (x 2 ) ˆσ ( ab l l(1γ) ) q i (x, M 2 ) = 1 x O q γ (x, M 2 ) = q a (x) q a (x, M 2 ) q a (x, M 2 ) i=q, q dz q i ( x z, M 2) α 2π Q2 i + q γ ( x z, M 2) α 1 x 2π Q2 i dz q i ( x z, M 2) α 2π Q2 i [ ( ( M 2 P q qγ (z) log [ ( P γ q q (z) log m 2 i ) ( M 2 ))] m 2 q [ ( P q γq (z) log )] 2 log(1 z) 1 ( ) M 2 m 2 i + + f q (z) + f γ (z) (3.3) )] 2 log(1 z) f(z) generalization to the multiple emission case: in each emission the leading singularity is removed the integrated cross-section is independent of the initial state quark masses Check: Total cross-section for different values of the initial state quark masses (CC channel) Including exact O(α) corrections Best we can: O(α) matched with parton-shower M_up M_up / 5 M_up / ±.22 (pb) 253.9±.23 (pb) ±.24 (pb) M_up M_up / 5 M_up / ±.28 (pb) ±.32 (pb) ±.35 (pb) 16

19 MRST 24 QED and photon induced processes QED evolution photon density in the proton photon induced processes γu dµ + ν µ γ d γ ν u γ W + W + ν u µ + γ ν W + µ + d µ + Charged Current channel same perturbative order as the O(α) corrections they contribute to the inclusive DY cross section depending on the cut on the final state jet, important effect on the lepton transverse momentum distribution W µ + W ν u d u d µ γ µ µ γ µ µ µ Neutral Current channel also new lowest order partonic subprocess γ γ q γ q γ, Z l q l γ q l l l γ, Z q q γ, Z l γ l q l q γ q γ, Z q l l (a) (b) (c) (d) 17

20 W transverse mass distribution (peak region) dσ (pb/gev) dm Born exact O(α) best M W T = G_mu scheme M (GeV) 2 p l pν (1 cos φ lν) reconstructed in the transverse plane jacobian peak at the W mass rather insensitive to QCD initial state radiation (e.g. ptw modeling) δ (%) M (GeV) recombined electrons show partial KLN cancelation bare (i.e. perfectly isolated) muons receive large final state corrections insensitive to photon induced processes relevant for the extraction of MW G_mu scheme rec. e + µ + 18

21 dσ (pb) dηl W-rapidity and lepton pseudo-rapidity distributions Born exact O(α) µ + exact O(α) e + best µ + best e dσ (pb) dyw δ (%) Born best µ + best e µ + e y W η l y W relevant for acceptances, luminosity monitoring, pdfs constraining (flat) correction factor ranges from -2% (W) to -4% (lepton) of the same order of present NNLO-QCD uncertainty Anastasiou et al. tion of a differential distribution at NNLO in 19

22 Z observables: invariant mass distribution dσ dm (pb/gev) l + l dσ dm (pb/gev) l + l e-4 1e-5 1e-6 1e-7 1e-8 1e M l + l (GeV) Born O(α) O(α) + h.o M l + l (GeV) Born O(α) O(α) + h.o. huge radiative corrections below the Z peak (final state radiation) in the large mass tail, large negative corrections (EW Sudakov logs) not negligible effect of (tree-level) photon-induced subprocess δ (%) δ (%) h.o. [µ] h.o. [e] M l + l (GeV) O(α) [µ] O(α) + γ-ind. [µ] O(α) [e] h.o. [µ] h.o. [e] M l + l (GeV) O(α) [µ] O(α) + γ-ind. [µ] O(α) [e] 2

23 Z observables: transverse mass and Z rapidity distributions dσ (pb/gev) dm dσ (pb) dyz M (GeV) Born O(α) O(α) + h.o. Born O(α) O(α) + h.o y Z above the Z peak, not negligible effect of the photon-induced processes Z rapidity: QED h.o. and photon-induced contribute at the per mille level δ (%) δ (%) h.o M (GeV) h.o O(α) O(α) + γ-ind y Z O(α) O(α) + γ-ind. 21

24 W/Z transverse mass ratio (preliminary) dσ/dx (pb) W Born W O(α) Z Born 5 Z O(α) 5 R(X) Born O(α) X ( dσ ) ( dσ R = dx W / orrection ) do not cancel!, X V = MV /M V dx Z the pqcd radiative corrections partially cancel in the ratio (Giele, Keller, Phys.Rev.D57:4433 (1998) ) the systematics due to the pdfs partially cancel in the ratio delicate discussion about the systematics on the acceptances the EW radiative corrections do not cancel in the ratio the ratio is very sensitive to the precise value of M W 22

25 The Drell-Yan processes and QCD dynamics at the LHC the lepton pair is (very often) accompanied by additional jets N= N=1 N=2 N=3 Tevatron, no cuts 92.1 % 7.6 %.3 % LHC, no cuts 79 % 15 % 5 %.1 % LHC, MT>1 TeV 3 % 38 % 21 % 8 % at LHC the cross section with N=1 enhanced by the subprocess with a gluon in the initial state (gluon density larger than at the Tevatron) the large MT cut forces the showering process ( enhances N=1,2,3) xf Q = 1 GeV ZEUS-JETS (prel.) 94- total error.6 H1 PDF 2 exp. error model error xu v.4 xg (.5) xd v.2 xs (.5) x 23

26 Combining QCD and EW corrections First attempt: combination of soft-gluon resummation with final state QED corrections Q.-H. Cao and C.-P. Yuan, Phys. Rev. Lett. 93 (24) 421 ResBos-A Additive combination of QCD and EW corrections: [ ] { } {[ ] dσ dσ dσ = + do do do QCD EW QCD EW [ dσ do ] Born } HERW IG P S QCD = ALPGEN (with CKKM-MLM Parton Shower matching), ResBos-CSS, MC@NLO, FEWZ, MCFM EW = HORACE interfaced with HERWIG QCD Parton Shower NLO-EW corrections convoluted with QCD PS inclusion of not reliable when hard non collinear radiation is important O(αα s ) terms Beyond the additive approximation, a full 2-loop O(αα s ) calculation is needed see: J.H. Kühn, A.Kulesza, S.Pozzorini, M.Schulze, hep-ph/73283 W. Hollik, T.Kasprzik, B.A. Kniehl, arxiv:

27 Monte Carlo tuning: Tevatron and LHC Monte Carlo ALPGEN FEWZ HORACE ResBos-A σ LO (pb) 96.3(3) 96.2(16) 95.64(4) 95.26(24) Table: MC tuning at the Tevatron for the LO cross section of the process p p W ± µ ± ν µ, using CTEQ6M with µ R = µ F = x 1 x 2 s Monte Carlo ALPGEN FEWZ HORACE σ LO (pb) 831(2) 834(2) 837.9(2) Table: MC tuning at the LHC for the LO cross section of the process pp W ± µ ± ν µ, using MRST24QED with µ R = µ F = p 2,W + M W 2 Monte Carlo σnlo Tevatron (pb) σnlo(pb) LHC MC@NLO (4) 2939(19) FEWZ 2643.(8) 211(14) Table: MC tuning for MC@NLO and FEWZ NLO inclusive cross sections of the process pp ( ) W ± µ ± ν µ, with CTEQ6M (Tevatron) and MRST24QED (LHC) After appropriate tuning, and with same input parameters and cuts, Monte Carlos agree at.1% level (or better) Alessandro Vicini - University of Milano Padova, April 26th 27 25

28 the LHC: M W and p µ distributions (pb) dσ dm W δ(%) LO HORACE + HERWIG PS MC@NLO MC@NLO+HORACE QCD EW EW+QCD (GeV) M W (pb) dσ dp µ δ(%) LO HORACE + HERWIG PS MC@NLO MC@NLO+HORACE 3 QCD 25 EW 2 EW+QCD p µ (GeV) the relative effect expressed in units Born+PS positive QCD corrections compensate negative EW corrections around the jacobian peak EW corrections mandatory to extract only QCD-Parton Shower is not sufficient the convolution with QCD Parton Shower modifies the relative effect and shape of the EW corrections M W 26

29 the LHC: M W and p µ distributions M W > 1 TeV, p µ > 5 GeV which relation between large negative EW Sudakov logs and QCD corrections? (fb) dσ dm W δ(%) e-4 1e LO HORACE + HERWIG PS MC@NLO MC@NLO+HORACE (GeV) M W QCD EW EW+QCD (fb) dσ dp µ the relative effect expressed in units Born+PS δ(%) LO HORACE + HERWIG PS MC@NLO MC@NLO+HORACE QCD EW EW+QCD p µ (GeV) negative QCD corrections sum up with negative EW corrections the sum 4( 7)% for M W 1.5(3) TeV and 3( 5)% for p µ.5(1) TeV 27

30 Conclusions the event generator HORACE contains almost the state of the art of EW corrections to CC and NC Drell-Yan processes a detailed phenomenological analysis demonstrates the impact of the EW corrections on several distributions and, in turn, on the measurement of several observables acceptances : pdfs, luminosity transverse mass : measurement of (limits on Higgs, MSSM) M W a realistic description of the Drell-Yan processes requires the combination of QCD and EW corrections (possibly in a unified generator) the interplay of the two sets of corrections is not trivial the QCD-Parton Shower provides the correct lowest order approximation of the kinematics of these processes and modifies the impact of the EW corrections the ratio of W/Z M_T distributions to extract the W mass is under study in the long term: unify ALPGEN and HORACE in a single generator 28

31 Back-up slides Alessandro Vicini - University of Milano Padova, April 26th 27 29

32 Electroweak results with HORACE LHC energy: S=14 TeV pdf: MRST24QED process: pp µ ± ν + X input scheme: selection cuts: α(), M W, M Z p,l and p,ν > 25GeV, η l < 2.5 extra cuts in photon-induced processes: p,jet < 3GeV, η jet > 2.5 3