QCD@LHC 2019 Measurement of the strong coupling constant by CMS Juska Pekkanen on behalf of the CMS Collaboration juska@cern.ch SUNY Buffalo July 16, 2019
The strong coupling constant α S � Not a constant but running wildly • Leads to confinement and asymptotic freedom of QCD � The only free parameter in QCD • Considering quark masses fixed � Enters every QCD theory calculation • Contributes e.g. to Higgs production σ calculation ⋆ Affects also BSM searches looking for deviations in Higgs σ ! • Precision measurement important but hard • δ ( α EM ) = 0.2 parts per billion, δ ( α S ) = 10 7 ppb ( ∼ 1%) � Determined in lattice QCD or collider experiments • Best precision from lattice calculations, but reliability of uncertainty estimates under discussions QCD@LHC 2019 Juska Pekkanen, juska@cern.ch July 16, 2019 2/14
History of α S measurements Strong coupling constant value vs year 0.125 0.120 World average α S ( m Z ) 0.115 0.110 0.105 0.100 1985 1990 1995 2000 2005 2010 2015 2020 Year � First known world average value from G. Altarelli in 1989 • Heavily extrapolated to m Z , but still within 1 σ ! � Lattice calculations and collider experiment results combined in world averages • Uncertainty estimation and choosing results not trivial � Converging to α s ( m Z ) = 0.118, but uncertainty grown QCD@LHC 2019 Juska Pekkanen, juska@cern.ch July 16, 2019 3/14
Measurements from CMS � CMS has produced O (10) α s extractions with 7–13 TeV � Here I present four interesting and/or recent results � Arbitrary choice, apologies if your favourite not included � Following analyses made the cut this time: • α s extraction from tt cross section @ 7 TeV • α s extraction from jet substructure in tt events • α s extraction from triple-differential dijet cross section • α s extraction from tt cross section @ 13 TeV QCD@LHC 2019 Juska Pekkanen, juska@cern.ch July 16, 2019 4/14
α s from tt production cross section � Inclusive σ tt from a previous CMS analysis is compared to NNLO QCD predictions • 2.3fb − 1 @ 7 TeV, arXiv:1208.2671 � α s ( m Z ) determined by fixing m t (and vice versa ) • Favoured α s from a PDF fit χ 2 scan � Five different NNLO PDF sets tested � First α s ( m Z ) result with NNLO at a hadron collider and the first α s extraction using top-quark production QCD@LHC 2019 Juska Pekkanen, juska@cern.ch July 16, 2019 5/14
Results PDG 2018 � Extracted α s values mostly below the 2012 world average • ... but world average came down from that! � With the latest world average results ∼ within 1 σ � Main result from NNPDF2.3: α s ( m Z ) = 0.1151 + 0 . 0028 − 0 . 0027 QCD@LHC 2019 Juska Pekkanen, juska@cern.ch July 16, 2019 6/14
α s from triple-differential dijet cross section � Triple differential σ dijet measured from dijet events with 20fb − 1 of 8 TeV data • σ dijet ( p ave T , rapidity separation y ∗ , boost y b ) � Dijet processes sensitive to α s � PDF fits done with the measeured σ dijet and DIS data from HERA • α s ( m Z ) inferred by repeating PDF fit with it as a free parameter QCD@LHC 2019 Juska Pekkanen, juska@cern.ch July 16, 2019 7/14
19.7 fb 1 (8 TeV) 10 8 Results dp T, avg dy b dy * [pb /GeV] y * < 1 (×10 2 ) 0 y b < 1 0 CMS 10 7 y * < 2 (×10 2 ) 0 y b < 1 1 y * < 3 (×10 1 ) 0 y b < 1 2 y * < 1 (×10 1 ) 10 6 1 y b < 2 0 y * < 2 (×10 1 ) 1 y b < 2 1 10 5 y * < 1 (×10 0 ) 2 y b < 3 0 10 4 NLOJET++ (NLO EW NP) NNPDF 3.0 10 3 = p T, max e 0.3 y * d 3 anti k t R = 0.7 10 2 10 1 10 0 10 1 10 2 10 3 10 4 200 300 500 1000 p T, avg [GeV] � Measured triple-differential σ dijet agree well with the NLOJ ET ++ and NNPDF 3.0 prediction • Some deviation in the boosted region y b > 1 � Extracted strong coupling: α s ( m Z ) = 0.1199 + 0 . 0034 − 0 . 0025 • Theory scale uncertainty dominating: + 0 . 0031 − 0 . 0020 QCD@LHC 2019 Juska Pekkanen, juska@cern.ch July 16, 2019 8/14
α s from jet substructure data in tt events � Measurement of jet substructure in tt → ℓ ± + jets events � s = 13 TeV • Using the full 2016 dataset; 36 fb − 1 at � Jet substructure sensitive to the strenght of QCD • Here angle between subjets most sensitive � α s ( m Z ) extracted at LO+LL theory accuracy • Measured distributions compared to POWHEG + PYTHIA 8 predictions, ’preferred’ α s determined in χ 2 scan l + b ν W + t t W – q b q' QCD@LHC 2019 Juska Pekkanen, juska@cern.ch July 16, 2019 9/14
Results -1 -1 35.9 fb (13 TeV) 35.9 fb (13 TeV) 2 2 χ χ CMS CMS 700 λ 0 700 ∆ (N) R g 0 POWHEG+PYTHIA 8 POWHEG+PYTHIA 8 ε bottom jets bottom jets Uncertainties 600 z 600 g charged particles Experimental charged particles ∆ R g Model + α 0.015 (m ) = 0.115 LO+LL, 2-loop CMW 500 500 FSR scale − S 0.013 Z Total LO+LL, 2-loop CMW 400 400 300 300 200 200 100 100 0 0 0.08 0.09 0.1 0.11 0.12 0.13 0.14 0.08 0.09 0.1 0.11 0.12 0.13 0.14 α α FSR (m ) FSR (m ) S Z S Z � Out of many jet substructure variables, the angle between groomed subjets ∆ R g gives the best α s extraction result � α s is determined from a FSR-process � Result: α s = 0.115 + 0 . 015 − 0 . 013 • Precision limited by FSR scale unc. in PYTHIA 8, + 0 . 014 − 0 . 012 QCD@LHC 2019 Juska Pekkanen, juska@cern.ch July 16, 2019 10/14
α s from tt production cross section (revisited) � Inclusive σ tt is measured from dileptonic tt events • Full 2016 dataset; 36 fb − 1 @ 13 TeV � α s ( m Z ) determined by fixing m t (and vice versa ) and comparing obtained σ tt to NNLO prediction • Preferred α s value from a χ 2 scan � Four different NNLO PDF sets tested QCD@LHC 2019 Juska Pekkanen, juska@cern.ch July 16, 2019 11/14
Results CMS 35.9 fb -1 (13 TeV) CMS 35.9 fb -1 (13 TeV) 10 χ 2 2018 PDG σ tt _ NNPDF3.1 CT14 MMHT14nnlo 7.5 m t (m t ) = 163.47 GeV MMHT14 ABMP16 CT14nnlo m t (m t ) = 163.30 GeV 5 NNPDF3.1nnlo m t (m t ) = 162.56 GeV 2.5 ABMP16nnlo m t (m t ) = 160.86 GeV 0 0.105 0.11 0.115 0.12 0.125 0.105 0.11 0.115 0.12 α S (m Z ) α S (m Z ) � Good agreement between all four NNLO PDF sets � Extracted α s values (again) below the world average � Just to pick one, NNPDF3.1 gives α s ( m Z ) = 0.1140 + 0 . 0039 − 0 . 0033 QCD@LHC 2019 Juska Pekkanen, juska@cern.ch July 16, 2019 12/14
Summary 0.130 0.125 0.120 0.115 α S ( m Z ) 0.110 0.105 World average 0.100 CMS Measurement with total uncertainty CMS Measurement without scale uncertainty 0.095 1990 1995 2000 2005 2010 2015 2020 Year � CMS results well in line with the world average � Uncertainties much larger than in combined values • Theory scale unc. dominate in σ dijet & jet substructure � Similar unc. with 2.3fb − 1 7 TeV and 36fb − 1 13 TeV in σ tt • Syst. limited, leading uncertainties from σ tt and PDF QCD@LHC 2019 Juska Pekkanen, juska@cern.ch July 16, 2019 13/14
References & further reading Links to the presented CMS results: 1. "Determination of the top-quark pole mass and strong coupling constant from the � s = 7 TeV" arXiv:1307.1907 tt production cross section in pp collisions at 2. "Measurement of the triple-differential dijet cross section in proton-proton � s = 8 TeV and constraints on parton distribution functions" collisions at arXiv:1705.02628 3. "Measurement of jet substructure observables in tt events from proton-proton � s = 13TeV" arXiv:1808.07340v2 collisions at 4. "Measurement of the tt production cross section, the top quark mass, and the � s = 13 TeV" strong coupling constant using dilepton events in pp collisions at arXiv:1812.10505 Excellent review on α s measurements by G. Dissertori: � "The Determination of the Strong Coupling Constant" 1506.05407 A contribution to: "The Standard Theory up to the Higgs discovery - 60 years of CERN", L. Maiani and G. Rolandi (editors) QCD@LHC 2019 Juska Pekkanen, juska@cern.ch July 16, 2019 14/14
Recommend
More recommend
