Measurements of W and Z Production at \sqrts=13 TeV with the CMS Experiment at the LHC
NNuclear and Particle Physics Proceedings 00 (2021) 1–6
Nuclear andParticle PhysicsProceedings
Measurements of W and Z Production at √ s =
13 TeV with the CMSExperiment at the LHC ∗ B. Bilin Fonds National de la Recherche Scientifique, Universit´e Libre de Bruxelles (FNRS-ULB / IIHE), Brussels, Belgium
Abstract
This note presents selected measurements of W and Z boson production, carried out with the CMS experimentat the LHC, based on samples of events collected during 2015-2018 physics runs. W boson events were selectedcontaining an isolated, energetic electron or muon, while Z boson events were selected containing a pair of isolated,energetic electrons or muons. Presented results include searches for rare decays of W bosons to pions.
Keywords:
LHC, CMS, Standard Model, Vector Boson, 13 TeV, QCD-2020
1. Introduction
Processes involving W & Z boson production are oneof the best understood processes at hadron colliders.Leptonically decaying W and Z boson processes pro-vide an almost background-free environment, which al-lows probing various QCD e ff ects by studying kinemat-ics precisely. Fig. 1 shows a generic transverse mo-mentum ( p T ) distribution of Z bosons. By studyingthe p T spectrum, regions of phase space described byperturbative QCD as well as regions sensitive to non-perturbative e ff ects can be probed.By studying W & Z boson production, severalelectro-weak parameters can be measured precisely. Us-ing the data collected by the CMS experiment [1] intwo collision periods of the LHC(2010-2012 and 2015-2018), rare Standard Model (SM) processes can be stud-ied, such as rare decays of the bosons as well as theirproduction in Double Parton Scattering (DPS) process.In this note, several CMS results are highlighted, car-ried out using Run-II p-p collision data at √ s =
13 TeVcollected in 2015-2018 physics runs. ∗ Talk given at 23rd International Conference in Quantum Chro-modynamics (QCD 2020), 27 - 30 october 2020, Montpellier - FR
Email address: [email protected] (B. Bilin) On behalf of the CMS Collaboration Figure 1: Generic p T spectrum of Drell-Yan process with regions sep-arately sensitive to di ff erent QCD e ff ects.
2. Measurements of inclusive and di ff erential Z bo-son cross section Using 2016 p-p collision data corresponding to an in-tegrated luminosity of 35.9 fb − , CMS has measured [2]Z boson production cross section inclusively as well asdi ff erentially with respect to p T , ϕ ∗ η and y of di-leptonpairs (di-electrons di-muons) in the fiducial phase-space a r X i v : . [ h e p - e x ] J a n Nuclear and Particle Physics Proceedings 00 (2021) 1–6 requiring leptons with p T >
25 GeV, η < . < M ll <
106 GeV. The results are corrected for detec-tor e ff ects with an unfolding procedure. Table 1 presentsthe obtained inclusive cross-section results in the de-scribed fiducial phase-space. Table 1: Inclusive fiducial cross sections results in the dimuon anddielectron final states as well as the combination of the two [2].
Cross section σ B [pb] σ Z → µµ ± ±
17 (lumi) σ Z → ee ±
10 (syst) ±
18 (lumi) σ Z → ‘‘ ± ±
17 (lumi)
The di ff erential cross section measurements are pre-sented in absolute values as well as normalized to theinclusive cross section. By measuring the normalizedcross sections, several systematic uncertainties cancelout, providing ultimate precision, the uncertainties be-ing smaller than 0.4 % for ϕ ∗ η < . p T (Z) < ff erential cross sectionin p T (Z). Figs. 3 and 4 show the ratio of predictionsobtained with several state-of-the-art generators alsoinvolving dedicated treatment of the non-perturbativeQCD e ff ects to the measured results. As seen in Figs. 3and 4 the higher p T region is well described by calcu-lations including higher orders of QCD in the MatrixElement, whereas the cross section at low p T is bettermodelled by calculations including resummation tech-nique, as well as ones involving a TMD approach.
3. Measurements of Drell-Yan di ff erential cross sec-tion The CMS Collaboration has measured the Drell-Yan (DY) di ff erential cross section over a wide massrange [3] using 2015 p-p dataset corresponding to 2.8(2.3) fb − in the di-muon (di-electron) final states. Thetotal and fiducial cross section measurements are car-ried out di ff erentially to the mass of di-lepton pairs ina range of 15 < M ll < p T >
25 GeV and η < .
4. The fiducial and ab-solute cross section results are presented separately forthe di-electron and di-muon channels separately as wellas the combination of the two channels. The results arepresented using an unfolding technique to correct fordetector and acceptance e ff ects. Fig. 5 shows the di-muon invariant mass distribution before correcting fordetector e ff ects. Fig. 6 shows the measured cross sec-tion combining the two channels, obtained in full phase-space. The results are also corrected for e ff ects of QED [GeV] ZT p [ pb / G e V ] Z T / dp s d DataMINLOaMC@NLOPOWHEG
CMS - e + , e - m + m fi * g Z/ > 25 GeV T | < 2.4, p h | (13 TeV) -1 Figure 2: The measured absolute di ff erential cross sections with re-spect to p T (Z) for the combination of dimuon and dielectron finalstates. The shaded bands around the data points (black) representtotal experimental uncertainty. Results are compared to the predic-tions with MadGraph5 aMC@NLO (square red markers), POWHEG(green triangles), and POWHEG-MINLO (blue circles), where the er-ror bars around the predictions correspond to the combined statistical,PDF, and scale uncertainties [2]. [GeV] ZT p M I N L O / D a t a P O W H E G / D a t a a M C @ N L O / D a t a CMS ZT dp s d > 25 GeV T | < 2.4, p h | - e + , e - m + m fi * g Z/ (13 TeV) -1 Figure 3: The ratios of the predictions with Mad-Graph5 aMC@NLO (square red markers), POWHEG (greentriangles), and POWHEG-MINLO (blue circles) to the measurementsin bins of p T (Z) for the combination of dimuon and dielectron finalstates. The shaded bands around the data points (black) correspondto the total experimental uncertainty. The error bars around thepredictions correspond to the combined statistical, PDF, and scaleuncertainties [2]. FSR radiation, which a ff ect mainly the region below theZ peak. The measurements provide a good agreementwith the theory predictions from FEWZ generator cal-culating the inclusive DY production at NNLO in QCD. Nuclear and Particle Physics Proceedings 00 (2021) 1–6 [GeV] ZT p G ene v a / D a t a R e s bo s / D a t a PB T M D / D a t a CMS ZT dp s d > 25 GeV T | < 2.4, p h | - e + , e - m + m fi * g Z/ (13 TeV) -1 Figure 4: The ratios of the predictions with PB TMD (square redmarkers), RESBOS (green triangles), and GENEVA (blue circles) tothe measurements in bins of p T (Z) for the combination of dimuonand dielectron final states. The shaded bands around the data points(black) correspond to the total experimental uncertainty. The errorbars around the predictions correspond to the combined statistical,PDF, and scale uncertainties (only statistical uncertainty shown forRESBOS) [2]. E n t r i e s pe r b i n -
10 110 Data mm fi Z * / g Wt+tW+ttEWMisid.
CMS (13 TeV) -1 m [GeV]
20 30 100 200 1000 2000 D a t a / ( D Y + B k g ) Figure 5: Observed di-muon invariant mass spectrum within the de-tector acceptance, ”EW” representing di-boson processes and Drell-Yan to τ + τ − , ”Misid.” representing W + jets and QCD multijet back-ground contributions. Each MC contribution is normalized using themost accurate theoretical cross section value available [3].
4. Measurements of W boson rapidity, helicity,double-di ff erential cross sections, and chargeasymmetry CMS has measured [4] di ff erential cross section andcharge asymmetry for inclusive W boson production ) [GeV] mm m ( / d m [ pb / G e V ] s d - - - - - - -
10 110 Data
FEWZ (NNLO QCD + NLO EW)
CMS ) (13 TeV) mm ( -1 (ee) 2.8 fb -1 - m + m , - e + e fi */Z g m [GeV]
20 30 100 200 1000 2000 T heo r y / D a t a Stat. unc. Tot. unc. Theo. unc. (FEWZ)
Figure 6: The di ff erential DY cross section measured in the full phasespace for the two channels combined and as predicted by FEWZ atNNLO. The ratio between the data and the theoretical prediction ispresented in the bottom panel. The coloured boxes represent the the-oretical uncertainties [3]. using 2016 p-p collision data corresponding to 35.9fb − . The measurement is carried out using template fit-ting technique, for the two transverse polarization statesof W bosons. The di ff erential absolute cross sectionas well as its value normalized to the total inclusiveW boson production cross section are measured. Themeasurements are carried out over the rapidity range | y (W) | < .
5. Fig.7 shows the observed data events in p T ( µ ) and | η ( µ ) | unrolled bins for W + → µ + ν , overlayedwith signal and background processes obtained from thetemplate fit. In Fig. 8 normalized W + → l + ν crosssection is presented for left- and right-handed helicitystates, combining electron and muon channels. ) bin T ,p h Unrolled muon ( · E v en t s CMS (13 TeV) -1 GeV[26,27] GeV[27,28] GeV[28,29] GeV[29,30] GeV[30,31] GeV[31,32] GeV[32,33] GeV[33,34] GeV[34,35] GeV[35,36] GeV[36,37] GeV[37,38] GeV[38,39] GeV[39,40] GeV[40,41] GeV[41,42] GeV[42,43] GeV[43,44] GeV[44,45]
Data L+ W R+ W W Drell-Yan QCD nt fi W t quark Diboson ) bin T ,p h Unrolled muon ( D a t a / p r ed . Figure 7: Distribution of unrolled bin for W + → µ + ν events for ob-served data and signal plus background events, where the signal andbackground processes are normalized to the result of the template fit.The cyan band over the data-to-prediction ratio represents the uncer-tainty in the total yield in each bin after the profiling process [4]. Nuclear and Particle Physics Proceedings 00 (2021) 1–6 t o t s | / W / d | y s d (fit) L W (MC@NLO) L W (MC@NLO*) L W (fit) R W (MC@NLO) R W (MC@NLO*) R W D a t a s / T heo r y s CMS (13 TeV) -1 n + l fi + W | W |y Figure 8: Measured normalized W + → l + ν cross sectionwith respect to | y W | for the left-handed and right-handed helicitystates, electron and muon channels combined, compared to Mad-Graph5 aMC@NLO predictions. MadGraph5 aMC@NLO ∗ predic-tions are obtained after p T (W) weighting applied. The lightly-filledband corresponds to the expected uncertainty from the PDF variations, µ F and µ R scales, and α S [4]. In addition, the W boson double-di ff erential crosssection (d σ / dp T ( l )d | η ( l ) | ) and W charge asymmetry asa function of p T ( l ) and | η ( l ) | are measured, as shown inFig. 9. The measurements are also used to constrain theparton distribution functions using NNPDF3.0 set. C ha r ge a sy mm e t r y Measured MadGraph5_aMC@NLO S a ¯ PDFs n l fi W G e V [ , ] G e V [ , ] G e V [ , . ] G e V [ , ] G e V [ , . ] G e V [ , ] G e V [ , . ] G e V [ , ] G e V [ , . ] G e V [ , ] G e V [ , . ] G e V [ , ] G e V [ , . ] G e V [ , ] G e V [ , ] G e V [ , ] G e V [ , ] G e V [ , ] Measured MadGraph5_aMC@NLO S a ¯ PDFs n l fi W CMS (13 TeV) -1
50 100 150 200 250 300 [0.0, 2.4] ˛ | h | bin: | h Unrolled dressed lepton | - - O b s . - e x p . Figure 9: Double-di ff erential W boson charge asymmetry as a func-tion of p T ( l ) and | η ( l ) | unrolled over | η ( l ) | , compared to predictionsfrom MadGraph5 aMC@NLO(coloured bands) [4].
5. Searches for W boson rare decay modes to pions
CMS has carried out searches for exclusive decaysof W bosons to 3 π [5] using 2016 and 2017 datasetsof 77.3 fb − and to πγ [6] using 2016-2018 datasets of137 fb − . W ± → π ± π ± π ∓ search utilizes di- τ triggersand reconstructs π decays using hadronic τ algorithm, whereas W ± → π ± γ search deploys a novel techniquelooking for the exclusive decay mode in top quark pairevents, where one of the W’s decaying from the a topquark is further decaying to a lepton and neutrino (elec-tron or muon) and used to preselect the events, and theother W to π ± γ . Fig. 10 shows the observed m(3 π ) and E v en t s / G e V E v en t s / G e V (GeV) p m
50 100 150 200 250 O b s / E x p E v en t s / G e V Observed QCD multijet ll fi Z OthersUncertainty ) -6 (B = 10 p fi W (13 TeV) -1 CMS (GeV) gp m E v en t s Data ) -4 Signal (BR=10Drell-YanQCD+Jets g tt gn l fig W g - l + l fig ZttOthersMC uncert.Fit (13 TeV) -1
137 fb
CMS
Preliminary
50 55 60 65 70 75 80 85 90 95 100 (GeV) gp m - D a t a / M C Figure 10: Observed and expected distributions of m(3 π ) (top) and ofm( πγ ) for the sum of the lepton channels (bottom). Signal predictionsare normalised to B ( W → π ) = − (top) and B ( W → πγ ) = − (bottom) [5, 6]. m( πγ ) distributions compared with expected signal andbackgrounds. Signal predictions are normalised to B (W → π ) = − (top) and B ( W → πγ ) = − (bot- Nuclear and Particle Physics Proceedings 00 (2021) 1–6 tom).Upper limits at 95% CL for branching fractions havebeen set: B (W → π ) < . × − , B (W → πγ ) < . × − . Currently there is no theoretical calcula-tion of B ( W → π ) and obtained results, improvingthe existing limits, motivate the calculation of it. W → πγ results demonstrate a novel search technique forrare hadronic decays of W bosons at the LHC.
6. Evidence for DPS production of W boson pairproduction
CMS has carried out a search [7] for W bosonpair production from DPS processes, using same-signelectron-muon and di-muon pairs, using 2016 and 2017datasets corresponding to 77.4 fb − . In Fig. 11 the il-lustrations for WW production from DPS and SPS pro-cesses are shown separately. W ± q ( p q ( p ⌫` ± W ± q ( p q ( p ⌫` ± qq ⌫⌫q W ± W ± Z ` ± W ± W ± ` ± q qq ⌫q q ⌫ W ± ` ± q gq W ± ` ± Figure 11: Illustrations of DPS (left) and SPS (middle and right)W ± W ± production where both W’s decay leptonically [7]. The events are required to have two same-sign lep-tons with p T ( l , l ) > ,
20 GeV, | η (e) | < . | η ( µ ) | < . p Tmiss >
15 GeV. To suppress the SPS process,a requirement in number of jets is applied, N jets < p T (jet) >
30 GeV and | η (jet) | < .
5. Furthermore,events with a b-tagged jet with p T (bjet) >
25 GeV and | η (bjet) | < . µ or τ h candidates are also rejected. A BDT classifier is trainedto extract the signal, as shown in Fig. 12 for µ − µ − finalstate.DPS WW cross section is measured for the first timewith an observed significance of 3.9 standard devia-tions, as summarised in Table 2. Table 2: Observed cross section values for inclusive DPS WW pro-duction, obtained from the maximum likelihood fit to the final classi-fier distribution [7].
Value Significance(standard deviations) σ PYTHIA
DPSWW,exp σ factorizedDPSWW,exp σ DPSWW,obs ± ± σ eff + − mb — ) - m - m Bin number ( D a t a / b k g . Total background uncertainty – W – DPS W E v en t s Data WZ NonpromptRare * g W – W – DPS WZZ
CMS (13 TeV) - - m - m Figure 12: Distribution of the final BDT classifier output for the µµ final state, in the negative charge configuration [7].
7. Conclusion
CMS has a rich SM physics program covering vari-ous measurements, a selection of which is summarizedin this note including W and Z bosons using p-p colli-sion data at √ s =
13 TeV. The presented results providestringent tests of our models based on SM, probing per-turbative and non-perturbative QCD e ff ects, and provid-ing valuable input to improve theoretical models. References [1] S. Chatrchyan, et al., The CMS Experiment at the CERN LHC,JINST 3 (2008) S08004. doi:10.1088/1748-0221/3/08/S08004 .[2] A. M. Sirunyan, et al., Measurements of di ff erential Z bosonproduction cross sections in proton-proton collisions at √ s =
13 TeV, JHEP 12 (2019) 061. arXiv:1909.04133 , doi:10.1007/JHEP12(2019)061 .[3] A. M. Sirunyan, et al., Measurement of the di ff erential Drell-Yan cross section in proton-proton collisions at √ s =
13 TeV,JHEP 12 (2019) 059. arXiv:1812.10529 , doi:10.1007/JHEP12(2019)059 .[4] A. M. Sirunyan, et al., Measurements of the W boson rapidity,helicity, double-di ff erential cross sections, and charge asymmetryin pp collisions at √ s =
13 TeV arXiv:2008.04174 .[5] A. M. Sirunyan, et al., Search for W boson decays to threecharged pions, Phys. Rev. Lett. 122 (15) (2019) 151802. arXiv:1901.11201 , doi:10.1103/PhysRevLett.122.151802 .[6] CMS Collaboration, Search for the rare exclusive hadronicdecay of a W boson into a pion and a photon inproton-proton collisions at 13 TeV, CMS PAS-SMP-20-008,https: // cds.cern.ch / record / √ s =
13 TeV,
Nuclear and Particle Physics Proceedings 00 (2021) 1–6 arXiv:1909.06265 , doi:10.1140/epjc/s10052-019-7541-6doi:10.1140/epjc/s10052-019-7541-6