QCD Studies With CMS Data for Measurements of the Top-Quark Mass and the Strong Coupling Constant

Transcription

1 QCD Studies With CMS Data for Measurements of the Top-Quark Mass and the Strong Coupling Constant Sebastian Naumann-Emme DESY DPG-Frühjahrstagung, Mainz,

2 The Compact Muon Solenoid & Quantum Chromodynamics Candidate for a t t event with b J/ψ + X There is QCD in all these hadron collisions CMS 2011 / / 20 fb 1 of pp data at s = 7 / 8 TeV Thorough understanding and accurate modeling of QCD needed: for SM precision and for BSM searches see K. Lipka s talk this morning , Mainz S. Naumann-Emme: QCD Studies With CMS Data 2 / 25

3 Outline In the following, will discuss two topics: 1. The top-quark mass QCD modeling Pole mass via the t t cross section 2. The strong coupling constant From the t t cross section From differential jet cross sections Results are labeled with CMS publication IDs: SMP2-028 TOP3-007 etc. Further details on the web: 1/α S n f =5 CMS trijet/dijet CMS inclusive jets 2m top n f =6 3 Q [GeV] n f =12 (MSSM) , Mainz S. Naumann-Emme: QCD Studies With CMS Data 3 / 25

4 17 1 CDF RunII, l+jets L int L int L int D0 RunII, l+jets L int D0 RunII, di-lepton L int ATLAS 2011, l+jets L int ATLAS 2011, di-lepton L int CMS 2011, l+jets L int CMS 2011, di-lepton L int CMS 2011, all jets L int = 8.7 fb CDF RunII, di-lepton = 5.6 fb CDF RunII, all jets L int = 5.8 fb miss CDF RunII, E +jets T = 8.7 fb = 3.6 fb = 5.3 fb = 4.7 fb = 4.7 fb = 4.9 fb = 4.9 fb = 3.5 fb The Top-Quark Mass Tevatron+LHC m top combination - March 2014, L = 3.5 fb fb int ATLAS + CDF + CMS + D0 Preliminary ± 1.12 (0.52 ± 0.49 ± 0.86) ± 3.69 ( ± 2.01(1.43 ± 0.95 ± 1.04) ± 1.85 (1.26 ± 1.05 ± 0.86) ± 1.50 (0.83 ± 0.47 ± 1.16) ± 2.79 (2.36 ± 0.55 ± 1.38) ± 1.55 (0.23 ± 0.72 ± 1.35) ± 1.63 ( ± 1.06 (0.27 ± 0.33 ± 0.97) ± 1.52 ( ± 1.41(0.69 ± 3.13) ± 1.50) ± 1.46) ± 1.23) χ / ndf =4.3/ World comb ± 0.76 (0.27 ± 0.24 ± 0.67) χ 2 prob.=93% Tevatron March 2013 (Run I+II) ± 0.87 (0.51 ± 0.36 ± 0.61) Previous Comb. LHC September ± 0.95 (0.23 ± 0.26 ± 0.88) total (stat. ijes syst.) TOP m top [GeV] As G. Cortiana pointed out few minutes ago, a yet more precise knowledge and understanding of m t is needed for EW precision fits, the fate of the universe , Mainz S. Naumann-Emme: QCD Studies With CMS Data 4 / 25

5 The Top-Quark Mass World average: m t = ± 0.36 (stat+jes) ± 0.67 (syst) GeV CMS latest measurement, presented at Moriond this morning: m t = ± 0.19 (stat+jes) ± 0.75 (syst) GeV Lepton+jets channel, kinematic fittt correct and 2D ideogram method using mw rec tt wrong and mfit t JSF CMS Preliminary, 19.7 fb, Permutations / 5 GeV s = 8 TeV, l+jets Data/MC contour m t [GeV] JSF = ± (stat) ± (syst) 1σ 000 2σ 3σ CMS Preliminary, 19.7 fb contour contour tt unmatched 1, Permutations / 5 GeV Data/MC s = 8 TeV, l+jets Z+Jets W+Jets single top Data CMS Preliminary, 19.7 fb correct TOP tt tt tt wrong reco m W [GeV] unmatched 1, Permutations / 5 GeV Data/MC s = 8 TeV, l+jets Z+Jets W+Jets single top Data reco m W [GeV] CMS Preliminary, 19.7 fb tt correct tt tt wrong unmatched TOP , s = 8 TeV, l+jets Z+Jets W+Jets single top Data fit m t [GeV] , Mainz S. Naumann-Emme: QCD Studies With CMS Data 5 / 25

6 The Top-Quark Mass World average: m t = ± 0.36 (stat+jes) ± 0.67 (syst) GeV CMS latest measurement, presented at Moriond this morning: m t = ± 0.19 (stat+jes) ± 0.75 (syst) GeV Lepton+jets channel, kinematic fit and 2D ideogram method using mw rec and mfit t JSF CMS Preliminary, 19.7 fb, m t [GeV] JSF = ± (stat) ± (syst) s = 8 TeV, l+jets 1σ 2σ 3σ contour contour contour TOP4-001 TOP3-014 δm t (GeV) Fit calibration 0. p T - and η-dependent JES 0.18 Jet energy resolution 0.26 Missing transverse momentum 0.09 Lepton energy scale 0.03 b-tagging 0.02 Pileup 0.27 Non-t t background 0.11 Flavor-dependent JES 0.41 b fragmentation 0.06 Semileptonic B-hadron decays 0.16 Parton distribution functions 0.09 QCD scales (µ R, µ F ) 0.13 ME-PS matching threshold 0.15 ME generator 0.23 Underlying event 0.17 Color reconnection effects , Mainz S. Naumann-Emme: QCD Studies With CMS Data 6 / 25

7 Modeling a t t Event Hard-scattering matrix element (ME): MadGraph, Powheg, MC@NLO , Mainz S. Naumann-Emme: QCD Studies With CMS Data 7 / 25

8 Modeling a t t Event Decay (with spin correlations) Parton distribution functions (PDFs): CTEQ, MSTW, NNPDF , Mainz S. Naumann-Emme: QCD Studies With CMS Data 8 / 25

9 Modeling a t t Event Multi-parton interactions (MPI) Initial-state radiation (ISR) Parton shower (PS) and hadronization: Pythia (Lund string fragmentation) or Herwig (cluster fragmentation) , Mainz S. Naumann-Emme: QCD Studies With CMS Data 9 / 25

10 Modeling a t t Event Color reconnections (CR) , Mainz S. Naumann-Emme: QCD Studies With CMS Data / 25

11 Differential Top-Mass Studies Check dependence of m t measurement on event kinematics, comparing to different ME generators, hadronization models, UE tunes probing in particular variables that could be sensitive to color-reconnection effects = 8 TeV, l+jets contour 1σ JSF contour 2σ contour s 1 CMS Preliminary, 19.7 fb, 3σ TOP m [GeV] t 2D > [GeV] <m t 2D m t [GeV] data MG Z2* 3 Data Difference to our standard MC: MadGraph+Pythia6 with Z2* tune 1 CMS Preliminary, 19.7 fb, MG, Pythia Z2* Powheg, Pythia Z2* s = 8 TeV, l+jets MG, Pythia P11 MG, Pythia P11noCR MC@NLO, Herwig η b,had , Mainz S. Naumann-Emme: QCD Studies With CMS Data 11 / 25

12 Differential Top-Mass Studies (cont.) TOP D <m t > [GeV] data MG Z2* [GeV] m 2D t CMS Preliminary, 19.7 fb Data MG, Pythia Z2* Powheg, Pythia Z2* 1, s = 8 TeV, l+jets MG, Pythia P11 MG, Pythia P11noCR MC@NLO, Herwig p [GeV] T,t,had 2D > [GeV] <m t 2D m t [GeV] data MG Z2* CMS Preliminary, 19.7 fb Data MG, Pythia Z2* Powheg, Pythia Z2* 1, s = 8 TeV, l+jets MG, Pythia P11 MG, Pythia P11noCR MC@NLO, Herwig R qq Studied 14 variables summed χ 2 /Ndf = 37/47 (prob. = 85%) with MadGraph+Pythia6 and Z2* tune Within current statistics: data overall well modeled by MC With more data: Will be able to constrain model uncertainties further , Mainz S. Naumann-Emme: QCD Studies With CMS Data 12 / 25

13 <p > T Κ data/sim Underlying Event and Color Reconnection UE study, using high-purity t t sample in the dilepton channel: <p > T Κ data/sim <p <p > > T T Κ data/sim Κ <p > T Κ data/sim TOP3-007 Event-by-event axis: p T (t t) p T (b 1 ) + p T (b 2 ) + p T (l) + p T ( l) + p miss Subtract b 1,b 2,l, l to characterize the soft, underlying activity Look at data/mc ratios, e.g. for the average p T of charged particles 0.9 as a function of p T (t t) CMS preliminary, 19.7 fb, s=8 TeV [ =0 extra jets ] P11 MPI hi TEV No CR [ 2 jets ] [ =1 extra jet ] [ =1 extra jet ] tt transverse momentum [GeV] Default MC: MadGraph+Pythia6 with tune Z2* Here: comparison to different Perugia 11 tunes 1.2 Description 1.1 of data 20% worse when switching off color reconnections , 0.8 Mainz S. Naumann-Emme: QCD Studies With CMS Data 13 / 25 T

14 ) (GeV dn/dm µ + µ B Fragmentation and Exclusive Decays CMS data preliminary, fb, J/ψ ψ' s=8 TeV l+jets channel dilepton channel [GeV] M µ + - µ TOP GeV<m(µµ)< 3.2 GeV Clear signal of b J/ψ + X in t t events Overall rate and kinematic distributions reasonably well modeled Future measurement: m t via m(l, J/ψ) Will require more precise fragmentation studies with LHC data Events / 0.25 Events / Data CMS preliminary, 19.8 fb, (PY6)+BG min R (J/ψ - jet) CMS preliminary, 19.8 fb, s = 8 TeV s = 8 TeV tt tt tt l + jets channel (MG) others (MG) Single top Dibosons Z/γ* ll W lν Data (PY6)+BG p (J/ψ)/p (nearest jet) T T tt tt tt l + jets channel (MG) others (MG) , Mainz S. Naumann-Emme: QCD Studies With CMS Data 14 / 25 Single top Dibosons Z/γ* ll W lν

15 Conceptional Limitations of m t Measurements 1. Aren t you just measuring a parameter of your MC event generators? Practically all m t measurements need calibration to MC However, according to implementation: close to pole mass in the sense of Breit-Wigner mass peak m MC t In 2011, A. Buckley et al.: shift could be O(1 GeV) Discussion at TOP2013: m MC t = mt pole within MeV, uncertainty being mainly from: 2. Isn t the pole mass ill-defined for a quark (colored object) anyway? Intrinsic renormalon ambiguity: 270 GeV for mt Motivated to bring m t measurements to uncertainties of 0.5 GeV also at a hadron collider But: proper modeling of non-perturb. effects required... especially when reconstructing invariant mass of decay products , Mainz S. Naumann-Emme: QCD Studies With CMS Data 15 / 25

16 Mass Dependence of t t Cross Section Since last year: σ(pp t t + X ) calculated at NNLO+NNLL QCD Uncertainties from higher orders (µ R, µ F ), PDF, α S, and m t now 3% each Compare to CMS most precise result for σ t t: s = 7 TeV, dileptons, 4% uncertainty (pb) tt σ s = 7 TeV; α S (m Z ) = Tevatron 2012 CMS, L = 2.3 fb Top++ 2.0, ABM11 Top++ 2.0, CT Top++ 2.0, HERAPDF1.5 Top++ 2.0, MSTW2008 Top++ 2.0, NNPDF (GeV) pole m t TOP2-022 Due to acceptance corrections, meas. σ t t also depends on m t For blue curve assume: = mt pole ± 1 GeV m MC t Full α S -PDF correlations taken into account Only minor differences between 4 of 5 PDF sets; ABM has smaller gluon density , Mainz S. Naumann-Emme: QCD Studies With CMS Data 16 / 25

17 Top Mass Resulting from t t Cross Section Obtain most probable mass value from marginalized joint posterior with Bayesian confidence interval First NNLO determination of the top mass Results compatible with average from direct m t measurements But approach not competitive in precision ABM11 CT MSTW2008 With α S (m ) = ± Z With α S (m ) of PDF set Z HERAPDF1.5 NNPDF2.3 CMS, mt pole = GeV with NNPDF2.3 s = 7 TeV, L = 2.3 fb ; NNLO+NNLL for σ tt Tevatron 2012 TOP (GeV) pole m t (σmeas ) ±1.4 (PDF) ±0.9 (µ t t R,F ) ±0.7 (α S ) ±0.9 (E LHC ) ±0.5 (mt MC ) GeV , Mainz S. Naumann-Emme: QCD Studies With CMS Data 17 / 25

18 The Strong Coupling Constant Besides quark masses, α S is the only free parameter of QCD Lagrangian Renormalization Group Equation predicts its energy dependence Measured in variety of processes and at different energies, evolved to Q = m Z for comparison 2013 world average: α S (m Z ) = ± (based on NNLO and N 3 LO) 0.5% uncertainty, but still significant for precise calculations Tension between and within different categories (especially DIS structure funcs, but also e + e ) Precision driven by low-q data and lattice QCD τ-decays Lattice DIS e + e - annihilation Z pole fits α s (Μ Ζ) , Mainz S. Naumann-Emme: QCD Studies With CMS Data 18 / 25

19 (pb) tt σ pole s = 7 TeV; mt = GeV CMS, L = 2.3 fb Top++ 2.0, ABM11 Top++ 2.0, CT Top++ 2.0, HERAPDF1.5 Top++ 2.0, MSTW2008 Top++ 2.0, NNPDF2.3 α S from the t t Cross Section α S (m ) Z pole CMS, s = 7 TeV, L = 2.3 fb ; NNLO+NNLL for σ tt ; m t = ± 1.4 GeV Default α S (m ) of respective PDF set Z PDG PDG TOP2-022 Same approach as for the m t extraction, but now leave α S free Need external m t constraint: take Tevatron average, adding ± 1 GeV for mt MC? = mt pole First determination of α S at NNLO from a hadron collider ABM11 CT α S (m Z ) = with NNPDF2.3 HERAPDF1.5 MSTW2008 NNPDF α S (m ) Z Competitive precision (similar to results using LEP2 data) , Mainz S. Naumann-Emme: QCD Studies With CMS Data 19 / 25

20 Three different measurements: α S from Jets SMP2-028 Double-diff. inclusive jet cross section: d 2 σ dp T dy QCD1-003 Ratio of inclusive trijet/dijet cross sections: R 32 ( p T 1,2 ) SMP2-027 Double-diff. trijet cross section: d 2 σ dm 3 dy max d y (pb/gev) T σ/dp 2 d 13 4 y < 0.5 ( ) CMS < y < 1.0 ( s = 7 TeV 1.0 < y < 1.5 ( L = 5.0 fb anti-k T µ = µ = p R F T R = 0.7 NNPDF2.1 NP Corr. 1.5 < y < 2.0 ( 2.0 < y < 2.5 ( Jet p (GeV) T ) ) ) ) R 32 Data/Theory CMS s = 7 TeV anti-k T R = 0.7 Data (Int. Lumi. = 5.0 fb ) NLO NPC NNPDF2.1-NNLO α s (M ) = Z Scale uncertainty PDF uncertainty p (GeV) T1, p (GeV) T1,2 d 2 σ/dm3dymax [pb / GeV] CMS preliminary L = 5.0 fb 1 s = 7 TeV Theory y max 1 1 < y max Anti-kT R = 0.7 MSTW NLO m3 [GeV] , Mainz S. Naumann-Emme: QCD Studies With CMS Data 20 / 25

21 α S from Jets: Method All three analyses using the 7 TeV data and same methodology for α S : 1. Compare measured cross section to QCD prediction at NLO Using NLOJet++; adding MC-based non-perturbative corrections; for the inclusive jets additionally EW corrections 2. Use PDF sets provided for series of α S (m Z ) values to determine α S dependence preserves PDF-α S correlations Ratio to NLO(CT-NLO) NPC y <0.5 Data (Int. Lumi. = 5.0 fb ) CT α s (M ) = Min. Value Z CT α s (M ) = Z CT α s (M ) = Max. Value Z CMS Preliminary s = 7 TeV anti-k T R = Jet p T 3 2 (GeV) χ CMS Preliminary CT-NLO y <2.5 Polynomial Fit α S (M ) Z 3. Obtain best α S (m Z ) and uncertainty via χ 2 summed over bins , Mainz S. Naumann-Emme: QCD Studies With CMS Data 21 / 25

22 α S from Jets: Results α S (m Z ) δrel max (scale/total) probed Q TOP2-022 t t (NNLO) % / 2.9% m t = 173 GeV SMP2-028 incl. jets (NLO) % / 5.5% p T = TeV QCD1-003 trijet/dijet (NLO) % / 4.8% p T 1,2 = TeV SMP2-027 trijet mass (NLO) % / 6.2% m 3 /2 = TeV α S (Q) 0.24 CMS Preliminary CMS Incl. Jets : α )= S (M Z CMS R CMS tt cross section CMS 3-Jet mass 0.18 CMS Incl. Jets D0 inclusive jets D0 angular correlation H1 ZEUS Q (GeV) , Mainz S. Naumann-Emme: QCD Studies With CMS Data 22 / 25

23 R CMS s = 7 TeV anti-k T R = Probing the Running of α S Data (Int. Lumi. = 5.0 fb ) NNPDF2.1 α s (M ) = Min. Value Z NNPDF2.1 α s (M ) = Z QCD1-003 NNPDF2.1 α s (M ) = Max. Value Z p (GeV) T1,2 p T 1, GeV GeV GeV Q 474 GeV 664 GeV 896 GeV data points 6 5 α S (m Z ) ± ± ± prob(χ 2 ) 49% 21% 77% α S (Q) ± ± ± By construction, we obtain for any Q bin directly α S (m Z ) If assumed evolution not valid: α S (m Z ) shifted For illustration of running, can later evolve back: α S (Q) Note: RGE validity also assumed in employed PDFs Strength of R 32 ratio: reduction of exp. and theor. uncertainties and of dependence on PDF evolution , Mainz S. Naumann-Emme: QCD Studies With CMS Data 23 / 25

24 Running of α S at High Energies α S (Q) depends on number of active flavors at Q 15 ( ) 1-loop approximation: 1 α S (Q 2 ) = 1 α S (µ 2 ) + β 0 ln Q 2 µ 2 with β 0 = 33 2n f 12π 14 CMS trijet/dijet CMS inclusive jets 1/α S inner error bars: exp. uncertainty n f =5 2m top n f =6 3 Q [GeV] n f =12 (MSSM) 4 gauge-coupling unification at 16 GeV? Seeing / not seeing change of slope: evidence for / against new colored particles , Mainz S. Naumann-Emme: QCD Studies With CMS Data 24 / 25

25 Conclusions & Outlook Measuring m t requires excellent understanding of QCD effects and the LHC data already allow us to study these t t cross section enabled first α S at NNLO from a hadron collider Jets probing α S up to TeV scale already, but higher precision requires NNLO prediction Probing the running of α S at highest energies will require considerable work by theory and experiment Will we be able to measure the running of m t? Maybe: m MS t from differential t t cross section up to very high p T (t t) , Mainz S. Naumann-Emme: QCD Studies With CMS Data 25 / 25

26 BACKUP , Mainz S. Naumann-Emme: QCD Studies With CMS Data 26 / 25

27 QCD-Model Uncertainties of TOP4-001 Hadronization: Flavor-dependent JES: Difference in jet energies from Pythia6 (Lund string fragmentation) to Herwig++ (cluster fragmentation), evaluated for light quarks, b-quarks, and gluons separately, then added in quadrature b fragmentation: Retuned Bowler-Lund fragmentation in Pythia to describe x B data from ALEPH and DELPHI, difference to Z2* tune taken as systematic uncertainty Semileptonic B-hadron decays: Varied branching ratios within PDG uncertainties; direct impact on neutrino fractions , Mainz S. Naumann-Emme: QCD Studies With CMS Data 27 / 25

28 Hard scattering: QCD-Model Uncertainties of TOP4-001 Parton distribution functions: Difference between CTEQ6.6L and envelope of PDF+α S uncertainties from CT, MSTW2008, and NNPDF2.3 Renormalization and factorization scales: Q and 4 ME-PS matching threshold: Default value of 20 GeV varied to and 40 GeV ME generator: Difference between MadGraph (LO multileg) and Powheg (NLO); additionally difference between measured and predicted top p T spectrum Non-perturbative QCD: Underlying event: Comparison between Pythia tunes: P11 vs. P11mpiHi and P11TeV Color reconnections: Comparison between Pythia tunes: P11 vs. P11noCR , Mainz S. Naumann-Emme: QCD Studies With CMS Data 28 / 25

29 Compatibility of m t Result at 7 and 8 TeV Analyis at 7 TeV: TOP1-015 m t = ± 0.43 (stat+jes) ± 0.98 (syst) GeV δm t (GeV) Fit calibration 0.06 p T - and η-dependent JES 0.28 Jet energy resolution 0.23 Missing transverse momentum 0.06 Lepton energy scale 0.02 b-tagging 0.12 Pileup 0.07 Non-t t background 0.13 b-jes 0.61 Parton distribution functions 0.07 QCD scales (µ R, µ F ) 0.24 ME-PS matching threshold 0.18 Underlying event 0.15 Color reconnection effects 0.54 Analyis at 8 TeV: TOP4-001 m t = ± 0.19 (stat+jes) ± 0.75 (syst) GeV δm t (GeV) Fit calibration 0. p T - and η-dependent JES 0.18 Jet energy resolution 0.26 Missing transverse momentum 0.09 Lepton energy scale 0.03 b-tagging 0.02 Pileup 0.27 Non-t t background 0.11 Flavor-dependent JES 0.41 b fragmentation 0.06 Semileptonic B-hadron decays 0.16 Parton distribution functions 0.09 QCD scales (µ R, µ F ) 0.13 ME-PS matching threshold 0.15 ME generator 0.23 Underlying event 0.17 Color reconnection effects , Mainz S. Naumann-Emme: QCD Studies With CMS Data 29 / 25

30 Compatibility of m t Result at 7 and 8 TeV Analyis at 7 TeV: TOP1-015 Analyis at 8 TeV: TOP4-001 m t = ± 0.43 (stat+jes) ± 0.98 (syst) GeV m t = ± 0.19 (stat+jes) ± 0.75 (syst) GeV Shift from 7 to 8 TeV: 1.45 GeV Significance: δ 2 7TeV + δ2 8TeV 2 ρ δ 7TeV δ 8TeV Correlation ρ: 0 for (stat+jes) low for experimental uncertainties, due to new MC/data corrections and scale factors high for theory uncertainties, since no drastic change in s Realistic assumptions for ρ significance of shift: σ , Mainz S. Naumann-Emme: QCD Studies With CMS Data 30 / 25

31 Additional Jets in Top-Pair Events Differential t t cross section as a function of the jet multiplicity: CMS Preliminary, 19.6 fb at s = 8 TeV CMS Preliminary, 19.6 fb at s = 8 TeV dσ djets σ 1 1 jet Dilepton Combined p > 30 GeV T Data MadGraph+Pythia MC@NLO+Herwig POWHEG+Pythia TOP2-041 dσ djets σ 1 1 jet Dilepton Combined p > 30 GeV T Data MadGraph+Pythia 2 4*Q 2 Q /4 Matching up Matching down TOP Data/MC Jets Compare to different ME generators and hadronization models Data/MC Jets... variations of Q 2 scale and threshold for ME-PS matching , Mainz S. Naumann-Emme: QCD Studies With CMS Data 31 / 25

32 Constraining PDFs with Jets SMP2-028 Using same CMS inclusive-jet data and NLO theory as before, can also study impact on PDFs: 1.0 CMS Preliminary Q 2 =1.9 GeV 2 Fits performed with HERAFitter framework 0.8 HERA DIS + CMS Jets HERA DIS As baseline: HERA-I DIS data, similar to HERAPDF1.0 but with more flexible parametrization xf(x,q 2 ) sea g u V d V Fixed α S (m Z ) = Simultan. fit of PDF and α S with HERA+CMS data yields: α S (m Z ) = experimental+non-perturb. uncertainty only , Mainz S. Naumann-Emme: QCD Studies With CMS Data 32 / 25 x

33 1.0 CMS Preliminary Q 2 =1.9 GeV 2 HERA DIS + CMS Jets HERA DIS 0.8 Constraining PDFs with Jets SMP2-028 xf(x,q 2 ) sea g u V d V Most significant impact: gluon at medium high x Fract. Uncert. xf(x,q 2 ) x CMS Preliminary gluon, Q 2 =1.9GeV HERA DIS + CMS Jets 2.5 HERA DIS Q=1.4 GeV x x Fract. Uncert. CMS Preliminary 90 gluon, Q 2 =1 4 GeV 2 HERA DIS + CMS Jets 80 HERA DIS Q=0 GeV x , Mainz S. Naumann-Emme: QCD Studies With CMS Data 33 / 25 xf(x,q 2 ) x

34 A W (η l ) = Constraining PDFs with W Asymmetry SMP2-021 dσ (W dη + ) dσ (W l dη ) l dσ (W dη + )+ dσ (W l dη ) l Muon charge asymmetry in W production: sensitive to d v, u v, d/u Charge asymmetry a (a) p > 25 GeV T Data CMS, L = 4.7 fb at s = 7 TeV x d v CMS NLO 13 parameter fit Q 2 =1.9 GeV 2 HERA I DIS + CMS A W HERA I DIS 0.15 NLO FEWZ + NLO PDF, 68% CL CT NNPDF23 HERAPDF15 Ratio MSTW2008 MSTW2008CPdeut Muon η 0.5 (HERA I DIS + CMS A W ) / (HERA I DIS) x MSTW2008CPdeut: variant of MSTW2008 with more flexible parametrization and deuteron corrections; changes uv and dv Again HERAFitter analysis with HERA-I DIS data as baseline; NLO from MCFM for W data , Mainz S. Naumann-Emme: QCD Studies With CMS Data 34 / 25

35 Constraining PDFs with W+charm W+charm production directly sensitive to strangeness in the proton dσ(w+c)/d η [pb] CMS L = 5.0 fb at s = 7 TeV p T l T p jet > 25 GeV > 35 GeV Data MSTW08 CT NNPDF23 NNPDF23 coll W l ν (l = µ,e) SMP2-002 Stat. uncertainty Total uncertainty η l Strangeness in standard PDFs based on neutrino-scattering data x s R s = (s + s ) / (u + d ) HERA I DIS + CMS W production CMS NLO free-s fit exp. unc. model unc. parametrization unc. Q 2 = 1.9 GeV 2 SMP x Adding W+c data to A W and HERA-I DIS data enables free-s fit; result consistent with similar ATLAS analysis and recent NOMAD result , Mainz S. Naumann-Emme: QCD Studies With CMS Data 35 / 25

36 PDF Constraints from Top Quarks Measured t t cross sections start providing constraints on the gluon PDF With the coming data, electroweak single-top production can be used to probe the b PDF and the u/d ratio (proton valence content: u/d = 2) NLO prediction from MCFM CMS, L = 19.7 fb, ABM11 CT CTw HERAPDF MSTW2008 NNPDF 2.3 s = 8 TeV CMS 1.95 ± 0. (stat.) ± 0.19 (syst.) n f = 4 n f = 5 n f = 4 n f = 5 n f = 4 n f = 4 TOP R t -ch. = σ t -ch. (t)/σ t -ch. (t) t-channel production: Kinematic regime: 0.02 x 0.5 Complementary to measurements via W asymmetry, which probe x 0.1 at the LHC and x 0.3 at the Tevatron , Mainz S. Naumann-Emme: QCD Studies With CMS Data 36 / 25