Prospects for Electroweak Measurements at the LHC
aa r X i v : . [ h e p - e x ] O c t th International Conference on High Energy Physics, Philadelphia, 2008
Prospects for Electroweak Measurements at the LHC
Martin W. Gr ¨unewald (on behalf of the ATLAS and CMS collaborations)
University College Dublin and University of Ghent
The prospects for electroweak measurements at the Large Hadron Collider (LHC) are discussed. In addition to high-luminosity results, special emphasis is placed on early start-up measurements with a total luminosity ranging from10/pb to 100/pb, using the general-purpose detectors ATLAS and CMS and their initially larger calibration andalignment uncertainties. Topics discussed here include inclusive W and Z production, W-boson mass, Z forward-backward asymmetry, Z-plus-jets production and di-boson production, the latter constraining trilinear electroweakgauge couplings. (Invited talk at the 34th ICHEP, Philadelphia, USA, July/August 2008)
1. INTRODUCTION
After more than a decade of construction, the LHC has successfully been turned on in September of 2008. Followingthe end-of-year shutdown, 2009 will see first proton-proton collisions, at centre-of-mass energies of perhaps 900 GeV(injection), 10 TeV and 14 TeV, with a total luminosity reaching 1/fb or more. This paper summarises the prospectsfor electroweak measurements at the LHC with the two general-purpose detectors ATLAS and CMS.Both ATLAS and CMS are built for high-pt physics at the LHC. Their most visible difference is in their size andmagnetic field arrangement: while ATLAS is dominated by a 2.5-3.5 Tesla toroid combined with a central 2 Teslasolenoid, CMS uses a large superconducting solenoid of 4 Tesla, with the outer return flux instrumented with muonchambers. The muon systems of the detectors provide a coverage of 2.7 (ATLAS) and 2.4 (CMS) in pseudo-rapidity | η | . The momentum resolution for muons of 1 TeV transverse momentum (pt) is 8% in ATLAS and 5% in CMS. Thecentral tracking system of both detectors covers | η | < .
5, providing resolutions of 3.7% (ATLAS) and 1.5% (CMS)for tracks with pt of 100 GeV. The electromagnetic crystal calorimeter of CMS provides a superior energy resolutionfor electromagnetic showers, 3%/ √ E ⊕ .
25% within | η | < .
0, versus 10% / √ E ⊕ .
5% within | η | < . | η | < .
9) provides for a better energy resolutioncompared to CMS ( | η | < . √ E ⊕
3% (ATLAS) and 70%/ √ E ⊕
8% (CMS) are expected. The resolutions vary strongly with η ; more details can be found in [1].Compared to the Tevatron centre-of-mass energy of 2 TeV, the factor 7 higher energy of the LHC results incross sections for the production of heavy particles to increase by up to one or two orders of magnitude. With aninstantaneous luminosity of 10 /cm /sec , the typical LHC event rates are 150 Hz W, 50 Hz Z and 1 Hz t ¯ t . In aninitial luminosity of 10/pb, we expect 150000 W → eν events, 15000 Z → ee events and 10000 t ¯ t events. Hence verysoon after LHC start-up, event statistics of electroweak final states will not limit the measurements.The signature of W → ℓν events is given by a high-energy lepton, missing (transverse) energy due to the neutrino,and a hadronic system recoiling against the decaying W boson. Inclusive Z → ℓ + ℓ − events contain a pair of high-energy leptons of opposite electric charge, no missing energy but again a hadronic recoil system. In both cases, thehadronic recoil spans the region from being very soft to hard, possibly even leading to one or more jets.
2. INCLUSIVE W/Z PRODUCTION
Inclusive W/Z production is considered as a fundamental benchmark process, also at the LHC, where it will bemeasured in the new regime of 10 TeV and 14 TeV center-of-mass energy. The cross section at 14 TeV is about180nb and 60nb for W and Z production, respectively, and is theoretically one of the best understood cross sectionsat hadron colliders, especially concerning uncertainties due to radiative corrections and parton distribution functions(PDFs). The W/Z process has the potential to become a high-pt reaction for the determination of the luminosity14 th International Conference on High Energy Physics, Philadelphia, 2008at the few % uncertainty level. Further, inclusive W/Z production serves as the starting point for more detailedanalyses, such as measuring the boson pt spectrum, looking for additional jets, or measurements of the Z-decayasymmetry and W-boson mass and width. In particular, Z events provide a crucial calibration source, given theprecise knowledge of the Z mass and width as measured at LEP.As an example, the ATLAS lepton identification and event selections designed for the LHC start-up, for 50/pb ofluminosity, are listed here [2]: electrons are identified within | η | < . | η | < .
5. Quality criteriaon electron candidates are assigned, such as “loose” (using the calorimetric showershape), “medium” (adding trackand track matching requirements), and “tight” (sharpening these requirements). For muon candidates, isolation, theamount of activity in a cone around the muon, is used.The selection of W → ℓν events requires either a medium-quality electron or an isolated muon. The transverseenergy of the lepton and the missing transverse energy must both be larger than 25 GeV, and the transverse massof the lepton-missing energy system must be larger than 40 GeV. In Z → ℓ + ℓ − , two charged leptons are present sothat the lepton identification is relaxed: either two loose isolated electrons or two isolated muons of opposite charge.In case of electrons, the transverse energy must be larger than 15 GeV and the invariant di-electron mass must bein the range from 80 to 100 GeV. In case of muons, the transverse momentum must be larger than 20 GeV, and thedi-muon mass be in a ±
20 GeV window around the Z mass value. The selection and trigger efficiencies range from60% to 90%. Both experiments exploit data-driven determinations using tag-and-probe on Z decays.The expected statistical and systematic uncertainties on the event numbers (rate) are as follows, using ATLAS50/pb, ATLAS 1/fb and CMS 1/fb selections [2, 3, 4]: For W events, the statistical uncertainties are 0.2%, 0.04% and0.04%, while the systematic uncertainties are projected as 3.1-5.2%, 2.4% and 3.3%. The systematic uncertainty isdominated by the missing energy determination. For the Z rate, the corresponding numbers are 0.8%, 0.2% and 0.13%statistical uncertainty, and 3.2-3.6%, 1.3% and 2.3% systematic uncertainty. The theoretical systematic uncertaintyis dominated by PDFs and the underlying boson pt distribution. Thus even with a small amount of luminosity atLHC start-up, the rate measurements are dominated by systematic uncertainties. The systematic uncertainties willdecrease with improved understanding of the detectors, but slower than the statistical uncertainty.In order to turn the rate into a cross section, a luminosity determination is needed, typically obtained by measuringforward scattering. The uncertainty on the luminosity from this method is estimated to be 10% initially, decreasingto about 5% in the long term. It is thus attractive to use W/Z production as an alternative luminosity reaction,because a smaller uncertainty can be achieved. Further, using a high-pt process similar to other signal processes,e.g., t ¯ t production, theoretical uncertainties due to PDFs and other issues partially cancel in the ratio. The mass of the W boson is a fundamental parameter of the electroweak Standard Model; in particular, togetherwith the mass of the top quark, it constrains the mass of the as yet undiscovered Higgs boson [5]. The W-boson massand width is measured precisely at LEP-2 [5] and by the Tevatron experiments CDF and DØ [6]. The measurementrequires a clean sample of W decays, thus tighter quality criteria on the lepton identification are imposed. In caseof ATLAS [7], one requires for W → eν exactly one isolated tight electron candidate, and for W → µν one isolatedmuon candidate. The transverse energy of the lepton and the missing transverse energy must both exceed 20 GeV.Already in 15/pb of luminosity, 67000 W → eν and 120000 W → µν events will be selected, together with 3000 Z → ee and 10000 Z → µµ events. The W mass will be extracted from the Jacobian peak observed in the transversemass of the lepton-neutrino system, or the transverse energy of the charged lepton. The Z events are a crucial sourceof calibration for the lepton energy scale (known Z mass) and energy resolution (known Z width), and used as well inthe determination of the differential lepton reconstruction efficiency. The low-luminosity ATLAS study shows howwell the energy scale and resolution can be monitored through Z events as a function of pseudo rapidity, for examplethe required corrections due to transition effects between central and endcap calorimeters. For 15/pb of luminosity,a W-mass uncertainty ranging from 160 to 240 MeV is expected. While this is not meant to be competitive withcurrent measurements at LEP-2 and the Tevatron, it serves to establish the W-mass analysis at the LHC. 24 th International Conference on High Energy Physics, Philadelphia, 2008Requiring higher luminosities, novel techniques were studied by the CMS collaboration to measure the W-bosonmass through templates generated from data (Z events), thus no longer relying on MC simulations [8]. Two possibil-ities are studied [9]: (i) an event-by-event transformation to change a Z event into a W event corresponding to a trialvalue of the W boson mass: one takes a Z → ℓℓ event, boosts it to the Z rest-system, rescales the lepton momenta bythe ratio M W ( trial ) /M Z ( LEP ), removes one lepton to mimic a neutrino, and boosts back to the detector system;and (ii) transformation of distributions, such as the lepton pt distribution [9]. The advantage of these methods liesin the fact that Z events from data rather than MC simulations are used, so that many systematic errors disappearand only the residual W-Z differences need to be studied. These methods require high luminosity as Z events fromdata are used. For 1/fb and 10/fb of luminosity, CMS expects statistical uncertainties of 40 and 15 MeV, withexperimental systematic errors of 40 and 20 MeV, and PDF uncertainties of 20 and 10 MeV, respectively.
Even in Z production in proton-proton collisions, a forward-backward asymmetry of the Z decay products is ex-pected. The Z is formed by a quark-antiquark pair; while the anti-quark always arises from the sea, the quark may alsobe a valence quark which on average carries a higher momentum than sea quarks. Thus the boost direction indicatesthe quark direction at high rapidities. In a sample of high-rapidity electron pairs, an asymmetry is observed whichcan be interpreted as a measurement of the effective electroweak mixing angle, sin θ eff , similar to the asymmetriesmeasured at LEP and at the Tevatron. With 100/fb of luminosity, ATLAS [10] expects a measurements of sin θ eff with a statistical precision of 0.00015 and a systematic uncertainty of 0.00024, comparable to the uncertainty of theworld average dominated by LEP and SLD [5]. The DØ experiment at the Tevatron has made a measurement witha statistical precision of 0.0018 and systematic uncertainty of 0.0006 using 1.1/fb [11]. The systematic uncertaintiesare by far dominated by PDF-related uncertainties, but the knowledge of PDFs is expected to improve throughmeasurements at the Tevatron, HERA, and also LHC (e.g., W asymmetry measurements).
3. Z PLUS JETS
Z production accompanied by jets serves as a test of perturbative QCD but is also a major background in searchesfor new physics, thus a good understanding of this process is required. The ATLAS [12] selection designed for aluminosity of 1/fb uses the standard Z → ee selection, while jets are clustered in a cone of R = 0 . | η | <
3. Jets are required to have a pt larger than 40 GeV. The lepton-jet separationmust exceed R = 0 .
4. The background from heavy-particle final states ( Z → τ τ , W , t ¯ t ) is taken from MC, whilethe QCD multi-jet background is derived from data, where MC simulations indicate that the expected multi-jetbackground fraction is independent of the jet pt. With 1/fb of luminosity, up to 4 jets can be observed in rate andpt spectrum, allowing to test MC models. A CMS study [13] specifically investigates Z + b ¯ b production resulting in asignature of two leptons and two b-jets. The background consists of Drell-Yan production plus light jets, Z + c ¯ c and t ¯ t production. The selection requires two isolated leptons of opposite charge and pt larger than 20 GeV, and at leasttwo b-tagged jets within | η | < . t ¯ t events the missing energy mustbe smaller than 50 GeV. Within 100/pb, this results in a cross section determination with a statistical (systematic)uncertainty of 15% (23%), the latter dominated by the jet energy scale and missing energy systematic.
4. DI-BOSON PRODUCTION
Pair-production of electroweak gauge bosons tests the triple gauge boson couplings of the electroweak StandardModel. Within the SM, the trilinear vertices
W W γ and
W W Z occur, while those involving only neutral gaugebosons, γ and Z, are absent. The charged triple gauge couplings (TGCs) are usually taken as g V , κ V and λ V for V = γ, Z ; they are related, for V = γ , to the magnetic dipole and electric quadrupole moment of the W boson.34 th International Conference on High Energy Physics, Philadelphia, 2008Within the SM, their values are g = κ = 1 and λ = 0. Di-boson production leads to final states containing chargedleptons from W/Z decay and phtons. The photon identification is similar to the electron identification except for aveto on charged tracks matching the calorimetric cluster of the photon candidate.CMS is using a cut-based analyses [8] while ATLAS studied in addition a boosted decision tree with improvedsensitivity compared to their cut-based analysis [14]. The number of selected events for signal (background) obtainedusing the ATLAS boosted-decision-tree selections on a total luminosity of 1/fb are: W γ : 3770 (2525) and Zγ W W : 469 (92) yielding a signal significance of10 standard deviations already with 0.1/fb of luminosity and a 20% background uncertainty;
W Z : 128 (16) yieldinga 5.8 sigma significance for 0.1/fb and a 20% background uncertainty; ZZ → ℓ ZZ → ℓ ν . ± .
5. CONCLUSION
The LHC will provide proton-proton collisions in 2009. The four detectors ATLAS, CMS, LHCb and ALICEare eagerly awaiting collision data. Both luminosity and cross sections at the LHC are much higher compared toearlier experiments, hence there will be no lack of statistics and sensitivity to rare processes such as ZZ production.The prospects of electroweak measurements are exciting due to high-performance detectors, allowing to place tightconstraints on the electroweak Standard Model, through measurements of production rates, masses and couplings ofthe electroweak gauge bosons. To exploit the data it is important to understand the early data and detectors quickly.
Acknowledgments
It is a pleasure to thank my colleagues from the CMS and ATLAS collaborations, notably Juan Alcaraz and TomLeCompte, for discussing results and answering patiently my questions.
References [1] The ATLAS Experiment at the CERN Large Hadron Collider, JINST 3 (2008) S08003;The CMS experiment at the CERN LHC, JINST 3 (2008) S08004.[2] ATLAS, Electroweak Boson Cross Section Measurements, in CERN-OPEN-2008-020, to appear.[3] CMS, Towards a Measurement of the Inclusive W → eν and γ ∗ /Z → ee Cross Sections in pp Collisions at √ s = 14 TeV, CMS PAS EWK-08-005.[4] CMS, Towards a Measurement of the Inclusive W → µν and γ ∗ /Z → µµ Cross Sections in pp Collisions at √ s = 14 TeV, CMS PAS EWK-07-002.[5] Peter Renton, Global Electroweak Fits and the Higgs Boson Mass, these proceedings.[6] Ashutosh Kotwal, Measurement of the W Mass, these proceedings.[7] ATLAS, Measurement of W boson mass at ATLAS with early data, in CERN-OPEN-2008-020, to appear.[8] CMS, CMS Technical Design Report, Vol.II: Physics Performance, J.Phys.G: Nucl. Part. Phys. 43 (2007) 995.[9] W.T.Giele and S.Keller, Determination of W boson properties at hadron colliders, hep-ph/9704419, Phys. Rev.D 57 (1998) 4433.[10] ATLAS, Forward-backward asymmetry in pp → ( Z/γ ∗ ) X → eeX Events, in CERN-OPEN-2008-020, to appear.[11] Susan Blessing, W and Z asymmetry measurements, these proceedings.[12] ATLAS, Measurement of Z+Jets Cross-Section at ATLAS, in CERN-OPEN-2008-020, to appear.[13] CMS, Measurement of the Zb ¯ b , Z → ℓℓℓℓ