ATLAS DETECTOR ANDPHYSICS PERFORMANCE

1

10

10 2

102

103

mH (GeV)

Sig

nal s

igni

fican

ce

H → γ γ + WH, ttH (H → γ γ ) WH, ttH (H → bb) H → ZZ(*) → 4 l

H → ZZ → llνν H → WW → lνjj

H → WW(*) → lνlν

Total significance

5 σ

100 fb-1

(no K-factors)

Technical Design Report

Issue: 1Revision: 0Reference: ATLAS TDR 15, CERN/LHCC 99-15Created: 25 May 1999Last modified: 25 May 1999Prepared By: ATLAS Collaboration

Volume II

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

All trademarks, copyright names and products referred to in this document are acknowledged as such.

For this edition typing and typograhical errors have been corrected. Layout and pagination may thereforediffer slightly with respect to the first, limited edition.

ii

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

ATLAS Collaboration

ArmeniaYerevan Physics Institute, Yerevan

AustraliaResearch Centre for High Energy Physics, Melbourne University, MelbourneUniversity of Sydney, Sydney

AustriaInstitut für Experimentalphysik der Leopold-Franzens-Universität Innsbruck, Innsbruck

Azerbaijan RepublicInstitute of Physics, Azerbaijan Academy of Science, Baku

Republic of BelarusInstitute of Physics of the Academy of Science of Belarus, MinskNational Centre of Particle and High Energy Physics, Minsk

BrazilUniversidade Federal do Rio de Janeiro, COPPE/EE/IF, Rio de Janeiro

CanadaUniversity of Alberta, EdmontonDepartment of Physics, University of British Columbia, VancouverUniversity of Carleton/C.R.P.P., CarletonGroup of Particle Physics, University of Montreal, MontrealDepartment of Physics, University of Toronto, TorontoTRIUMF, VancouverUniversity of Victoria, Victoria

CERNEuropean Laboratory for Particle Physics (CERN), Geneva

ChinaInstitute of High Energy Physics, Academia Sinica, Beijing, University of Science and Technology ofChina, Hefei, University of Nanjing and University of Shandong

Czech RepublicAcademy of Sciences of the Czech Republic, Institute of Physics and Institute ofComputer Science, PragueCharles University, Faculty of Mathematics and Physics, PragueCzech Technical University in Prague, Faculty of Nuclear Sciences andPhysical Engineering, Faculty of Mechanical Engineering, Prague

DenmarkNiels Bohr Institute, University of Copenhagen, Copenhagen

FinlandHelsinki Institute of Physics, Helsinki

FranceLaboratoire d’Annecy-le-Vieux de Physique des Particules (LAPP), IN2P3-CNRS, Annecy-le-VieuxUniversité Blaise Pascal, IN2P3-CNRS, Clermont-FerrandInstitut des Sciences Nucléaires de Grenoble, IN2P3-CNRS-Université Joseph Fourier, GrenobleCentre de Physique des Particules de Marseille, IN2P3-CNRS, MarseilleLaboratoire de l’Accélérateur Linéaire, IN2P3-CNRS, OrsayLPNHE, Universités de Paris VI et VII, IN2P3-CNRS, Paris

ATLAS Collaboration iii

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

CEA, DSM/DAPNIA, Centre d’Etudes de Saclay, Gif-sur-Yvette

Republic of GeorgiaInstitute of Physics of the Georgian Academy of Sciences and Tbilisi State University, Tbilisi

GermanyPhysikalisches Institut, Universität Bonn, BonnInstitut für Physik, Universität Dortmund, DortmundFakultät für Physik, Albert-Ludwigs-Universität, FreiburgInstitut für Hochenergiephysik der Universität Heidelberg, HeidelbergInstitut für Physik, Johannes-Gutenberg Universität Mainz, MainzLehrstuhl für Informatik V, Universität Mannheim, MannheimSektion Physik, Ludwig-Maximilian-Universität München, MünchenMax-Planck-Institut für Physik, MünchenFachbereich Physik, Universität Siegen, SiegenFachbereich Physik, Bergische Universität, Wuppertal

GreeceAthens National Technical University, AthensAthens University, AthensHigh Energy Physics Department and Department of Mechanical Engineering, Aristotle University ofThessaloniki, Thessaloniki

IsraelDepartment of Physics, Technion, HaifaRaymond and Beverly Sackler Faculty of Exact Sciences, School of Physics and Astronomy, Tel-AvivUniversity, Tel-AvivDepartment of Particle Physics, The Weizmann Institute of Science, Rehovot

ItalyDipartimento di Fisica dell’ Università della Calabria e I.N.F.N., CosenzaLaboratori Nazionali di Frascati dell’ I.N.F.N., FrascatiDipartimento di Fisica dell’ Università di Genova e I.N.F.N., GenovaDipartimento di Fisica dell’ Università di Lecce e I.N.F.N., LecceDipartimento di Fisica dell’ Università di Milano e I.N.F.N., MilanoDipartimento di Scienze Fisiche, Università di Napoli ‘Federico II’ e I.N.F.N., NapoliDipartimento di Fisica Nucleare e Teorica dell’ Università di Pavia e I.N.F.N., PaviaDipartimento di Fisica dell’ Università di Pisa e I.N.F.N., PisaDipartimento di Fisica dell’ Università di Roma ‘La Sapienza’ e I.N.F.N., RomaDipartimento di Fisica dell’ Università di Roma ‘Tor Vergata’ e I.N.F.N., RomaDipartimento di Fisica dell’ Università di Roma ‘Roma Tre’ e I.N.F.N., RomaDipartimento di Fisica dell’ Università di Udine, Gruppo collegato di Udine I.N.F.N. Trieste, Udine

JapanDepartment of Information Science, Fukui University, FukuiHiroshima Institute of Technology, HiroshimaDepartment of Physics, Hiroshima University, Higashi-HiroshimaKEK, High Energy Accelerator Research Organisation, TsukubaDepartment of Physics, Faculty of Science, Kobe University, KobeDepartment of Physics, Kyoto University, KyotoKyoto University of Education, Kyoto-shiDepartment of Electrical Engineering, Nagasaki Institute of Applied Science, NagasakiNaruto University of Education, Naruto-shiDepartment of Physics, Faculty of Science, Shinshu University, MatsumotoInternational Center for Elementary Particle Physics, University of Tokyo, TokyoPhysics Department, Tokyo Metropolitan University, TokyoDepartment of Applied Physics, Tokyo University of Agriculture and Technology, Tokyo

iv ATLAS Collaboration

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

MoroccoFaculté des Sciences Aïn Chock, Université Hassan II, Casablanca, and Université Mohamed V, Rabat

NetherlandsFOM - Institute SAF NIKHEF and University of Amsterdam/NIKHEF, AmsterdamUniversity of Nijmegen/NIKHEF, Nijmegen

NorwayUniversity of Bergen, BergenUniversity of Oslo, Oslo

PolandHenryk Niewodniczanski Institute of Nuclear Physics, CracowFaculty of Physics and Nuclear Techniques of the University of Mining and Metallurgy, Cracow

PortugalLaboratorio de Instrumentação e Física Experimental de Partículas (University of Lisboa, University ofCoimbra, University Católica-Figueira da Foz and University Nova de Lisboa), Lisbon

RomaniaInstitute of Atomic Physics, National Institute of Physics and Nuclear Engineering, Bucharest

RussiaInstitute for Theoretical and Experimental Physics (ITEP), MoscowP.N. Lebedev Institute of Physics, MoscowMoscow Engineering and Physics Institute (MEPhI), MoscowMoscow State University, Institute of Nuclear Physics, MoscowBudker Institute of Nuclear Physics (BINP), NovosibirskInstitute for High Energy Physics (IHEP), ProtvinoPetersburg Nuclear Physics Institute (PNPI), Gatchina, St. Petersburg

JINRJoint Institute for Nuclear Research, Dubna

Slovak RepublicBratislava University, Bratislava, and Institute of Experimental Physics of the Slovak Academy ofSciences, Kosice

SloveniaJozef Stefan Institute and Department of Physics, University of Ljubljana, Ljubljana

SpainInstitut de Física d’Altes Energies (IFAE), Universidad Autónoma de Barcelona, Bellaterra, BarcelonaPhysics Department, Universidad Autónoma de Madrid, MadridInstituto de Física Corpuscular (IFIC), Centro Mixto Universidad de Valencia - CSIC, Valencia

SwedenFysiska institutionen, Lunds universitet, LundRoyal Institute of Technology (KTH), StockholmUniversity of Stockholm, StockholmUppsala University, Department of Radiation Sciences, Uppsala

SwitzerlandLaboratory for High Energy Physics, University of Bern, BernSection de Physique, Université de Genève, Geneva

TurkeyDepartment of Physics, Ankara University, AnkaraDepartment of Physics, Bogaziçi University, Istanbul

ATLAS Collaboration v

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

United KingdomSchool of Physics and Astronomy, The University of Birmingham, BirminghamCavendish Laboratory, Cambridge University, CambridgeDepartment of Physics and Astronomy, University of Edinburgh, EdinburghDepartment of Physics and Astronomy, University of Glasgow, GlasgowDepartment of Physics, Lancaster University, LancasterDepartment of Physics, Oliver Lodge Laboratory, University of Liverpool, LiverpoolDepartment of Physics, Queen Mary and Westfield College, University of London, LondonDepartment of Physics, Royal Holloway and Bedford New College, University of London, EghamDepartment of Physics and Astronomy, University College London, LondonDepartment of Physics and Astronomy, University of Manchester, ManchesterDepartment of Physics, Oxford University, OxfordRutherford Appleton Laboratory, Chilton, DidcotDepartment of Physics, University of Sheffield, Sheffield

United States of AmericaState University of New York at Albany, New YorkArgonne National Laboratory, Argonne, IllinoisUniversity of Arizona, Tucson, ArizonaDepartment of Physics, The University of Texas at Arlington, Arlington, TexasLawrence Berkeley Laboratory and University of California, Berkeley, CaliforniaDepartment of Physics, Boston University, Boston, MassachusettsBrandeis University, Department of Physics, Waltham, MassachusettsBrookhaven National Laboratory (BNL), Upton, New YorkUniversity of Chicago, Enrico Fermi Institute, Chicago, IllinoisNevis Laboratory, Columbia University, Irvington, New YorkDepartment of Physics, Duke University, Durham, North CarolinaDepartment of Physics, Hampton University, VirginiaDepartment of Physics, Harvard University, Cambridge, MassachusettsIndiana University, Bloomington, IndianaUniversity of California, Irvine, CaliforniaMassachusetts Institute of Technology, Department of Physics, Cambridge, MassachusettsUniversity of Michigan, Department of Physics, Ann Arbor, MichiganMichigan State University, Department of Physics and Astronomy, East Lansing, MichiganUniversity of New Mexico, New Mexico Center for Particle Physics, AlbuquerquePhysics Department, Northern Illinois University, DeKalb, IllinoisOhio State University, Columbus, OhioDepartment of Physics and Astronomy, University of OklahomaDepartment of Physics, University of Pennsylvania, Philadelphia, PennsylvaniaUniversity of Pittsburgh, Pittsburgh, PennsylvaniaDepartment of Physics and Astronomy, University of Rochester, Rochester, New YorkInstitute for Particle Physics, University of California, Santa Cruz, CaliforniaDepartment of Physics, Southern Methodist University, Dallas, TexasState University of New York at Stony Brook, Stony Brook, New YorkTufts University, Medford, MassachusettsHigh Energy Physics, University of Illinois, Urbana, IllinoisDepartment of Physics, Department of Mechanical Engineering, University of Washington, Seattle,WashingtonDepartment of Physics, University of Wisconsin, Madison, Wisconsin

vi ATLAS Collaboration

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

Acknowledgements

The Editors would like to thank Mario Ruggier for his continuous help and competent adviceon all FrameMaker issues. The Editors also warmly thank Michèle Jouhet and Isabelle Canonfor the processing of the colour figures and the cover pages. Finally they would like to expresstheir gratitude to all the Print-shop staff for their expertise in printing this document.

Acknowledgements vii

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

viii Acknowledgements

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

Table Of Contents

14 Physics overview . . . . . . . . . . . . . . . . . . . . . . 45914.1 Introduction . . . . . . . . . . . . . . . . . . . . . . 45914.2 Theoretical picture . . . . . . . . . . . . . . . . . . . 46014.3 Challenges of new physics . . . . . . . . . . . . . . . . . 46214.4 Simulation of physics signals and backgrounds. . . . . . . . . . 463

14.4.1 Event generators. . . . . . . . . . . . . . . . . . 46414.4.2 Signal observability. . . . . . . . . . . . . . . . . 466

14.5 Outline . . . . . . . . . . . . . . . . . . . . . . . 46814.6 References . . . . . . . . . . . . . . . . . . . . . . 468

15 QCD processes at the LHC. . . . . . . . . . . . . . . . . . . 47115.1 Introduction . . . . . . . . . . . . . . . . . . . . . . 47115.2 Knowledge of the proton structure . . . . . . . . . . . . . . 472

15.2.1 Global parton analyses and parton kinematics at the LHC . . . 47215.2.2 Properties and uncertainties of parton distribution functions . . 47315.2.3 Expected improvements before the LHC start-up . . . . . . 47715.2.4 The role of data from ATLAS . . . . . . . . . . . . . 477

15.3 Properties of minimum−bias events. . . . . . . . . . . . . . 47815.3.1 Importance of minimum−bias studies . . . . . . . . . . 47815.3.2 Selection of minimum−bias events . . . . . . . . . . . 47815.3.3 Modelling of minimum-bias events . . . . . . . . . . . 47815.3.4 Measurements . . . . . . . . . . . . . . . . . . 479

15.4 Measurements of hard diffractive scattering . . . . . . . . . . . 48215.4.1 Overview . . . . . . . . . . . . . . . . . . . . 48215.4.2 Existing studies of hard diffraction . . . . . . . . . . . 48415.4.3 Models for hard diffractive scattering . . . . . . . . . . 48515.4.4 Trigger and event selection . . . . . . . . . . . . . . 48615.4.5 Single hard diffractive dissociation . . . . . . . . . . . 48815.4.6 Double Pomeron exchange . . . . . . . . . . . . . . 49115.4.7 Colour-singlet exchange . . . . . . . . . . . . . . . 49315.4.8 Diffractive W and Z production . . . . . . . . . . . . 49415.4.9 Diffractive heavy flavour production . . . . . . . . . . 49415.4.10 Summary on hard diffractive scattering . . . . . . . . . . 495

15.5 Jet physics . . . . . . . . . . . . . . . . . . . . . . 49615.5.1 Overview . . . . . . . . . . . . . . . . . . . . 49615.5.2 Inclusive jet cross-section . . . . . . . . . . . . . . . 49615.5.3 Jet shape and fragmentation . . . . . . . . . . . . . . 50115.5.4 Di-jet production . . . . . . . . . . . . . . . . . 50215.5.5 Multi-jet production . . . . . . . . . . . . . . . . 50615.5.6 Double parton scattering . . . . . . . . . . . . . . . 507

15.6 Photon physics . . . . . . . . . . . . . . . . . . . . . 50815.6.1 Overview . . . . . . . . . . . . . . . . . . . . 50815.6.2 Inclusive photon production . . . . . . . . . . . . . 50815.6.3 Photon pair production . . . . . . . . . . . . . . . 510

Table Of Contents ix

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

15.6.4 Photon + jet production . . . . . . . . . . . . . . . 51315.6.5 Photon + charm and photon + beauty production . . . . . . 514

15.7 Drell-Yan physics and gauge-boson production . . . . . . . . . 51515.7.1 Overview . . . . . . . . . . . . . . . . . . . . 51515.7.2 Drell-Yan production . . . . . . . . . . . . . . . . 51515.7.3 W production. . . . . . . . . . . . . . . . . . . 51715.7.4 Z production . . . . . . . . . . . . . . . . . . . 52115.7.5 Gauge boson pair production . . . . . . . . . . . . . 523

15.8 Heavy flavour physics . . . . . . . . . . . . . . . . . . 52715.8.1 Overview . . . . . . . . . . . . . . . . . . . . 52715.8.2 Charm production . . . . . . . . . . . . . . . . . 52815.8.3 Bottom production . . . . . . . . . . . . . . . . . 53015.8.4 Top production . . . . . . . . . . . . . . . . . . 534

15.9 Conclusion . . . . . . . . . . . . . . . . . . . . . . 53615.10 References . . . . . . . . . . . . . . . . . . . . . . 537

16 Physics of electroweak gauge bosons . . . . . . . . . . . . . . . 54516.1 Measurement of the W mass . . . . . . . . . . . . . . . . 545

16.1.1 The method . . . . . . . . . . . . . . . . . . . 54616.1.2 W production and selection . . . . . . . . . . . . . . 54716.1.3 Expected uncertainties . . . . . . . . . . . . . . . 54716.1.4 Results . . . . . . . . . . . . . . . . . . . . . 551

16.2 Gauge-boson pair production. . . . . . . . . . . . . . . . 55316.2.1 Wγ Production . . . . . . . . . . . . . . . . . . 55416.2.2 WZ Production . . . . . . . . . . . . . . . . . . 55516.2.3 Determination of Triple Gauge Couplings . . . . . . . . . 55516.2.4 Systematic uncertainties . . . . . . . . . . . . . . . 55716.2.5 Results . . . . . . . . . . . . . . . . . . . . . 558

16.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . 55916.4 References . . . . . . . . . . . . . . . . . . . . . . 559

17 B-physics . . . . . . . . . . . . . . . . . . . . . . . . . 56117.1 Introduction. . . . . . . . . . . . . . . . . . . . . . 561

17.1.1 General features of beauty production in ATLAS . . . . . . 56217.1.2 Model used for simulation studies . . . . . . . . . . . 56217.1.3 Trigger . . . . . . . . . . . . . . . . . . . . . 563

17.2 CP-violation studies . . . . . . . . . . . . . . . . . . . 56417.2.1 Overview . . . . . . . . . . . . . . . . . . . . 56417.2.2 Measurement of asymmetry in B0d → J/ψK0s . . . . . . . . 56517.2.3 Measurement of asymmetry in B0d → π+π- . . . . . . . . . . 57717.2.4 Analysis of the decay B0s → J/ψ φ . . . . . . . . . . . . 58217.2.5 Analysis of the decay B0d → D0K*0 . . . . . . . . . . . . . . 59017.2.6 Conclusions on CP violation . . . . . . . . . . . . . 591

17.3 Measurements of B0s oscillations . . . . . . . . . . . . . . . 59217.3.1 Introduction . . . . . . . . . . . . . . . . . . . 59217.3.2 Event reconstruction . . . . . . . . . . . . . . . . 593

x Table Of Contents

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

17.3.3 Background analysis . . . . . . . . . . . . . . . . 59717.3.4 Evaluation of signal and background statistics . . . . . . . 59717.3.5 Determination of the proper-time resolution . . . . . . . . 60017.3.6 Extraction of reach . . . . . . . . . . . . . . . . . 60117.3.7 Dependence of reach on experimental quantities . . . . . . 60317.3.8 Conclusions . . . . . . . . . . . . . . . . . . . 604

17.4 Rare decays B → µµ(X) . . . . . . . . . . . . . . . . . . 60417.4.1 Introduction . . . . . . . . . . . . . . . . . . . 60417.4.2 Theoretical approach . . . . . . . . . . . . . . . . 60517.4.3 Simulation of rare B-decay events . . . . . . . . . . . . 60617.4.4 The measurement of the forward–backward asymmetry . . . . 61017.4.5 Conclusions . . . . . . . . . . . . . . . . . . . 612

17.5 Precision measurements of B hadrons . . . . . . . . . . . . . 61217.5.1 Measurements with the Bc meson . . . . . . . . . . . . 61217.5.2 Λb polarisation measurement . . . . . . . . . . . . . 613

17.6 Conclusions on the B-physics potential . . . . . . . . . . . . 61517.7 References . . . . . . . . . . . . . . . . . . . . . . 616

18 Heavy quarks and leptons . . . . . . . . . . . . . . . . . . . 61918.1 Top quark physics . . . . . . . . . . . . . . . . . . . . 619

18.1.1 Introduction . . . . . . . . . . . . . . . . . . . 61918.1.2 tt selection and event yields . . . . . . . . . . . . . . 62018.1.3 Measurement of the top quark mass . . . . . . . . . . . 62218.1.4 Top quark pair production . . . . . . . . . . . . . . 63918.1.5 Top quark decays and couplings . . . . . . . . . . . . 64318.1.6 Electroweak single top quark production . . . . . . . . . 65218.1.7 Conclusions of top quark physics studies . . . . . . . . . 662

18.2 Fourth generation quarks . . . . . . . . . . . . . . . . . 66318.2.1 Fourth family up quarks . . . . . . . . . . . . . . . 66418.2.2 Fourth family down quarks . . . . . . . . . . . . . . 66618.2.3 Bound states of fourth family quarks. . . . . . . . . . . 667

18.3 Heavy leptons . . . . . . . . . . . . . . . . . . . . . 66818.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . 66918.5 References . . . . . . . . . . . . . . . . . . . . . . 669

19 Higgs Bosons . . . . . . . . . . . . . . . . . . . . . . . 67319.1 Introduction . . . . . . . . . . . . . . . . . . . . . . 67319.2 Standard Model Higgs boson . . . . . . . . . . . . . . . . 674

19.2.1 Introduction . . . . . . . . . . . . . . . . . . . 67419.2.2 H → γγ . . . . . . . . . . . . . . . . . . . . . 67519.2.3 H → Zγ . . . . . . . . . . . . . . . . . . . . . 68419.2.4 H → bb . . . . . . . . . . . . . . . . . . . . . 68519.2.5 H → ZZ* → 4l. . . . . . . . . . . . . . . . . . . 69319.2.6 H → WW(*) → lνlν . . . . . . . . . . . . . . . . . 70419.2.7 WH with H→ WW* → lνlν and W → lν . . . . . . . . . . 70919.2.8 Sensitivity to the SM Higgs boson in the intermediate mass range 712

Table Of Contents xi

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

19.2.9 H → ZZ → 4l . . . . . . . . . . . . . . . . . . . 71419.2.10 Heavy Higgs boson . . . . . . . . . . . . . . . . 71619.2.11 Overall sensitivity to the SM Higgs searches . . . . . . . . 72919.2.12 Determination of the SM Higgs-boson parameters . . . . . . 730

19.3 Minimal Supersymmetric Standard Model Higgs boson . . . . . . 73619.3.1 Introduction . . . . . . . . . . . . . . . . . . . 73619.3.2 Scenarios with heavy SUSY particles. . . . . . . . . . . 73719.3.3 Overall sensitivity . . . . . . . . . . . . . . . . . 77319.3.4 Determination of the MSSM Higgs parameters . . . . . . . 77719.3.5 SUGRA scenarios . . . . . . . . . . . . . . . . . 781

19.4 Strongly interacting Higgs sector . . . . . . . . . . . . . . 79519.4.1 Detector performance issues . . . . . . . . . . . . . 79519.4.2 Vector boson scattering in the Chiral Lagrangian model . . . . 796

19.5 Conclusions on the Higgs sector . . . . . . . . . . . . . . . 80119.6 References . . . . . . . . . . . . . . . . . . . . . . 803

20 Supersymmetry . . . . . . . . . . . . . . . . . . . . . . 81120.1 Introduction. . . . . . . . . . . . . . . . . . . . . . 81120.2 Supergravity models . . . . . . . . . . . . . . . . . . . 816

20.2.1 Inclusive SUGRA measurements . . . . . . . . . . . . 81820.2.2 Exclusive SUGRA measurements for moderate tan β . . . . . 82220.2.3 l+l- SUGRA signatures. . . . . . . . . . . . . . . . 82520.2.4 More complex leptonic SUGRA signatures . . . . . . . . 82920.2.5 h → bb SUGRA signatures . . . . . . . . . . . . . . 83520.2.6 Thresholds and model-independent SUGRA masses . . . . . 83920.2.7 Other signatures for SUGRA Points 1 – 5 . . . . . . . . . 84220.2.8 Exclusive SUGRA measurements for large tan β. . . . . . . 84720.2.9 Fitting minimal SUGRA parameters . . . . . . . . . . . 85320.2.10 Non-universal SUGRA models . . . . . . . . . . . . 860

20.3 Gauge mediated SUSY breaking models . . . . . . . . . . . . 86320.3.1 GMSB Point G1a . . . . . . . . . . . . . . . . . 86520.3.2 GMSB Point G1b . . . . . . . . . . . . . . . . . 87020.3.3 GMSB Point G2a . . . . . . . . . . . . . . . . . 87320.3.4 GMSB Point G2b . . . . . . . . . . . . . . . . . 87720.3.5 Fitting GMSB parameters . . . . . . . . . . . . . . 883

20.4 R-Parity breaking models . . . . . . . . . . . . . . . . . 88720.4.1 Baryon number violation: χ10 → qqq . . . . . . . . . . . 88820.4.2 Lepton number violation: χ10 → l+l-ν . . . . . . . . . . 89520.4.3 Lepton number violation: χ10 → qql, qqν . . . . . . . . . 906

20.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . 91020.6 References . . . . . . . . . . . . . . . . . . . . . . 911

21 Other physics beyond the Standard Model . . . . . . . . . . . . . 91521.1 Introduction. . . . . . . . . . . . . . . . . . . . . . 91521.2 Search for technicolor signals . . . . . . . . . . . . . . . . 915

21.2.1 Technicolor signals from qq fusion . . . . . . . . . . . 916

xii Table Of Contents

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

21.2.2 Signals from vector boson fusion . . . . . . . . . . . . 92421.2.3 Conclusion. . . . . . . . . . . . . . . . . . . . 925

21.3 Search for excited quarks . . . . . . . . . . . . . . . . . 92521.3.1 The widths of excited quarks . . . . . . . . . . . . . 92621.3.2 Simulation of the signal and backgrounds . . . . . . . . . 92721.3.3 Conclusion. . . . . . . . . . . . . . . . . . . . 930

21.4 Leptoquarks . . . . . . . . . . . . . . . . . . . . . . 93121.5 Compositeness . . . . . . . . . . . . . . . . . . . . . 932

21.5.1 High-pT jets . . . . . . . . . . . . . . . . . . . 93221.5.2 Transverse energy distributions of jets. . . . . . . . . . . 93321.5.3 Jet angular distributions. . . . . . . . . . . . . . . . 93521.5.4 Dilepton production . . . . . . . . . . . . . . . . 939

21.6 Search for new gauge bosons and Majorana neutrinos . . . . . . . 93921.6.1 Search for new vector bosons . . . . . . . . . . . . . 94021.6.2 Search for right-handed Majorana neutrinos . . . . . . . . 944

21.7 Monopoles . . . . . . . . . . . . . . . . . . . . . . 94921.8 References . . . . . . . . . . . . . . . . . . . . . . 952

A Members of the ATLAS Collaboration . . . . . . . . . . . . . . 955

Table Of Contents xiii

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

xiv Table Of Contents

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

14 Physics overview

14.1 Introduction

The ATLAS physics programme has been already discussed in several documents, the mostcomprehensive ones being the Letter of Intent [14-1] and the Technical Proposal [14-2]. Thegoals which have been defined there and which have guided the detector optimisation proce-dure remain essentially the same, the most important one being measurements that will lead toan understanding of the mechanism of electroweak symmetry breaking.

The high energy and luminosity of the LHC offers a large range of physics opportunities, fromthe precise measurement of the properties of known objects to the exploration of the high ener-gy frontier. The need to accommodate the very large spectrum of possible physics signatureshas guided the optimisation of the detector design. The desire to probe the origin of the elec-troweak scale leads to a major focus on the Higgs boson; ATLAS must be sensitive to it over thefull range of allowed masses. Other important goals are searches for other phenomena possiblyrelated to the symmetry breaking, such as particles predicted by supersymmetry or technicol-our theories, as well as new gauge bosons and evidence for composite quarks and leptons. Theinvestigation of CP violation in B decays and the precision measurements of W and top-quarkmasses and triple gauge boson couplings will also be important components of the ATLASphysics programme.

As discussed in the previous volume, and as also will be illustrated several times throughoutthis one, excellent performance of the detector is needed to achieve these physics goals.

• The various Higgs boson searches, which resent some of the most challenging signatures,were used as benchmark processes for the setting of parameters that describe the detectorperformance. High-resolution measurements of electrons, photons and muons, excellentsecondary vertex detection for τ−leptons and b-quarks, high-resolution calorimetry forjets and missing transverse energy (ETmiss) are essential to explore the full range of possi-ble Higgs boson masses.

• Searches for SUSY set the benchmarks on the hermeticity and ETmiss capability of the de-tector, as well as on b-tagging at high luminosity.

• Searches for new heavy gauge bosons provided benchmark requirements for high-resolu-tion lepton measurements and charge identification in the pT range as large as a few TeV.

• Signatures characteristic for quark compositeness set the requirements for the measure-ment of very high-pT jets.

• The precision measurements of the W and top-quark masses, gauge boson couplings, CPviolation and the determination of the Cabibbo-Kobayashi-Maskawa unitarity triangleyielded benchmarks that address the need to precisely control the energy scale for jetsand leptons, determine precisely secondary vertices, reconstruct fully final states with rel-atively low-pT particles and trigger on low-pT leptons.

14 Physics overview 459

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

14.2 Theoretical picture

The Standard Model (SM) [14-3] is a very successful description of the interactions of the com-ponents of matter at the smallest scales (10-18 m) and highest energies (~200 GeV) accessible tocurrent experiments. It is a quantum field theory which describes the interaction of spin-1/2point-like fermions, whose interactions are mediated by spin-1 gauge bosons. The bosons are aconsequence of local gauge invariance applied to the fermion fields and are a manifestation ofthe symmetry group of the theory, i.e. SU(3)xSU(2)xU(1) [14-3] [14-4].

The fundamental fermions are leptons and quarks. The left-handed states are doublets underthe SU(2) group, while the right-handed states are singlets. There are three generations of fermi-ons, each generation identical except for mass: the origin of this structure, and the breaking ofgenerational symmetry (flavour symmetry), remain a mystery. There are three leptons withelectric charge -1, the electron (e), muon (µ) and tau lepton (τ) and three electrically neutral lep-tons, the neutrinos νe, νµ and ντ. Similarly, there are three quarks with electric charge 2/3, up(u), charm (c) and top (t), and three with electric charge -1/3, down (d), strange (s) and bottom(b). The quarks are triplets under the SU(3) group and thus carry an additional ‘charge’, referredto as colour. There is mixing between the three generations of quarks, which is parametrised bythe Cabibbo-Kobayashi-Maskawa (CKM) [14-5] matrix whose origin is not explained by theStandard Model.

The SU(2)xU(1) symmetry group (which describes the so-called electroweak interaction) isspontaneously broken by the existence of a (postulated) Higgs field with non-zero expectationvalue [14-6]. This leads to the emergence of massive vector bosons, the W and Z, which mediatethe weak interaction, while the photon of electromagnetism remains massless. One physical de-gree of freedom remains in the Higgs sector, which should manifest as a neutral scalar bosonH0, which is presently unobserved. The SU(3) group describes the strong interaction (quantumchromodynamics or QCD) [14-4]. Eight vector gluons mediate this interaction. They carry col-our charges themselves, and are thus self-interacting. This implies that the QCD coupling αs issmall for large momentum transfers but large for small momentum transfers, and leads to theconfinement of quarks inside colour-neutral hadrons. Attempting to free a quark produces a jetof hadrons through production of quark-antiquark pairs and gluons.

The success of the SM of strong, weak and electromagnetic interactions has drawn increased at-tention to its limitations. In its simplest version, the model has 19 parameters, the three couplingconstants of the gauge theory SU(3)xSU(2)xU(1), three lepton and six quark masses, the mass ofthe Z boson which sets the scale of weak interactions, and the four parameters which describethe rotation from the weak to the mass eigenstates of the charge -1/3 quarks (CKM matrix). Allof these parameters are known with varying errors. Of the two remaining parameters, a CP-vio-lating parameter associated with the strong interactions must be very small. The last parameteris associated with the mechanism responsible for the breakdown of electroweak SU(2)xU(1) toU(1)em. This can be taken as the mass of the, as yet undiscovered, Higgs boson. The couplings ofthe Higgs boson are determined once its mass is given.

The gauge theory part of the SM has been well tested, but there is no direct evidence either foror against the simple Higgs mechanism for electroweak symmetry breaking. All masses are tiedto the mass scale of the Higgs sector. Although within the model there is no guidance about theHiggs mass itself, some constraints can be delivered from the perturbative calculations withinthe model requiring the Higgs couplings to remain finite and positive up to an energy scale Λ[14-7]. Such calculations exists at the two-loop level for both lower and upper Higgs massbounds. With present experimental results on the SM parameters, if the Higgs mass is in the

460 14 Physics overview

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

range 160 to 170 GeV [14-8] then the renormalisation-group behaviour of the Standard Model isperturbative and well behaved up to Planck scale ΛPl ~ 1019 GeV. For smaller or larger values ofmH new physics must set in below ΛPl.

As its mass increases, the self couplings and the couplings to the W and Z bosons grow [14-9].This feature has a very important consequence. Either the Higgs boson must have a mass lessthan about 800 GeV, or the dynamics of WW and ZZ interactions with centre-of-mass energiesof order 1 TeV will reveal new structure. It is this simple argument that sets the energy scale thatmust be reached to guarantee that an experiment will be able to provide information on the na-ture of electroweak symmetry breaking.

The presence of a single elementary scalar boson is unsatisfactory to many theorists. If the theo-ry is part of some more fundamental theory, which has some other larger mass scale (such as thescale of grand unification or the Planck scale), there is a serious ‘fine tuning’ or naturalnessproblem. Radiative corrections to the Higgs boson mass result in a value that is driven to thelarger scale unless some delicate cancellation is engineered ((m02 − m12) ~ mW2 where m0 and m1are order 1015 GeV or larger). There are two ways out of this problem which involve new phys-ics on the scale of 1 TeV. New strong dynamics could enter that provides the scale of mW, or newparticles could appear so that the larger scale is still possible, but the divergences are cancelledon a much smaller scale. In any of the options, Standard Model, new dynamics or cancellations,the energy scale is the same; something must be discovered at the TeV scale.

Supersymmetry [14-10] is an appealing concept for which there is so far no experimental evi-dence. It offers the only presently known mechanism for incorporating gravity into the quan-tum theory of particle interactions and provides an elegant cancellation mechanism for thedivergences, provided that at the electroweak scale the theory is supersymmetric. The successesof the Standard Model (such as precision electroweak predictions) are retained, while avoidingany fine tuning of the Higgs mass. Some supersymmetric models allow for the unification ofgauge couplings at a high scale and a consequent reduction of the number of arbitrary parame-ters.

Supersymmetric models postulate the existence of superpartners for all the presently observedparticles: bosonic superpartners of fermions (squarks and sleptons), and fermionic superpart-ners of bosons (gluinos and gauginos). There are also multiple Higgs bosons: h, H, A and H±.There is thus a large spectrum of presently unobserved particles, whose exact masses, couplingsand decay chains are calculable in the theory given certain parameters. Unfortunately these pa-rameters are unknown. Nonetheless, if supersymmetry is to have anything to do with elec-troweak symmetry breaking, the masses should be in the region below or order of 1 TeV.

An example of the strong coupling scenario is ‘technicolour’ for models based on dynamicalsymmetry breaking [14-11]. Again, if the dynamics is to have anything to do with electroweaksymmetry breaking we would expect new states in the region below 1 TeV; most models predicta large spectrum of such states. An elegant implementation of this appealing idea is lacking.However, all models predict structure in the WW scattering amplitude at around 1 TeV centre-of-mass energy.

There are also other possibilities for new physics that are not necessarily related to the scale ofelectroweak symmetry breaking. There could be new neutral or charged gauge bosons withmass larger than the Z and W; there could be new quarks, charged leptons or massive neutrinos,or quarks and leptons could turn out not to be elementary objects. While we have no definitiveexpectations for the masses of these objects, the LHC experiments must be able to search forthem over the available energy range.

14 Physics overview 461

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

Results on precision measurements within the Standard Model, as well as limits on new phys-ics, from present experiments are presented, case by case, in the relevant chapter of this volume.

14.3 Challenges of new physics

This volume presents examples of the physics programme which should be possible with theATLAS detector. The channels studied in previous documents [14-1][14-2] are re-examined andmany new strategies proposed.

In the initial phase at low luminosity, the experiment will function as a factory for QCD process-es, heavy flavour and gauge bosons production. This will allow a large number of precisionmeasurements in the early stages of the experiment.

A large variety of QCD related processes will be studied. These measurements are of impor-tance as studies of QCD ‘per se’ in a new energy regime with high statistics. Of particular inter-est will be jet and photon physics, open charm and beauty production and gauge bosonsproduction. A study of diffractive processes will present significant experimental challenges it-self, given the limited angular coverage of the ATLAS detector. Several aspects of diffractiveproduction of jets, gauge bosons, heavy flavour partons will be nevertheless studied in detail.LHC will extend the exploration of the hard partonic processes to large energy scales (of fewhundred GeV2), while reaching small fractional momentum of the proton being carried by ascattered partons (of 10-5). Precise constraints on the partonic distribution functions will be de-rived from measurements of Drell-Yan production, of W and Z bosons production, of produc-tion of direct photons and high-pT jets, heavy flavours and gauge boson pairs. Deviation fromthe theoretical predictions for QCD processes themselves might indicate the onset of new phys-ics, such as compositeness. Measurement and understanding of these QCD processes will be es-sential as they form the dominant background searches for new phenomena.

Even at low luminosity, LHC is a beauty factory with 1012 bb expected per year. The availablestatistics will be limited only by the rate at which data can be recorded. The proposed B-physicsprogramme is therefore very wide. Specific B-physics topics include the search for and meas-urement of CP violation, of Bs0 mixing and of rare decays. ATLAS can perform competitivehigh-accuracy measurements of Bs0 mixing, covering the statistically preferred range of theStandard Model predictions. Rare B mesons such as Bc will be copiously produced at LHC. Thestudy of B-baryon decay dynamics and spectroscopy of rare B hadrons will be also carried out.

LHC has a great potential for performing high precision top physics measurements with abouteight million tt pairs expected to be produced for an integrated luminosity of 10 fb-1. It wouldallow not only for the precise measurements of the top-quark mass (with a precision of ~2 GeV)but also for the detailed study of properties of the top-quark itself. The single top productionshould be observable and the high statistics will allow searches for many rare top decays. Theprecise knowledge of the top-quark mass places strong constraints on the mass of the StandardModel Higgs boson, while a detailed study of its properties may reveal as well new physics.

One of the challenges to the LHC experiments will be whether the precision of the W-massmeasurement can be improved. Given the 300 million single W events expected in one year ofdata taking, the expected statistical uncertainty will be about 2 MeV. The very ambitious goalfor both theory and experiment is to reduce the individual sources of systematic errors to less

462 14 Physics overview

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

than 10 MeV, which would allow for the measurement of the W mass with precision of betterthan 20 MeV. This would ensure that the precision of the W mass is not the dominant source oferrors in testing radiative corrections in the SM prediction for the Higgs mass.

The large rate of gauge boson pair production at the LHC enables ATLAS to provide criticaltests of the triple gauge-boson couplings. The gauge cancellations predicted by the StandardModel will be studied and measurements of possible anomalous couplings made. These probeunderlying non-standard physics. The most sensitive variables to compare with Standard Mod-el predictions are the transverse momentum spectra of high-pT photons or reconstructed Z bos-ons.

If the Higgs boson is not discovered before LHC begins operation, the searches for it and itspossible supersymmetric extensions in the Minimal Supersymmetric Standard Model (MSSM)will be a main focus of activity. Search strategies presented here explore a variety of possiblesignatures, being accessible already at low luminosity or only at design luminosity. Althoughthe cleanest one would lead to reconstruction of narrow mass peaks in the photonic or leptonicdecay channels, very promising are the signatures which lead to multi-jet or multi-τ final states.In several cases signal-to-background ratios much smaller than one are expected, and in mostcases detection of the Higgs boson will provide an experimental challenge. Nevertheless, theATLAS experiment alone will cover the full mass range up to 1 TeV for the SM Higgs and alsothe full parameter space for the MSSM Higgs scenarios. It has also a large potential for searchesin alternative scenarios.

Discovering SUSY at the LHC will be straightforward if it exists at the electroweak scale. Copi-ous production of squarks and gluinos can be expected, since the cross-section should be aslarge as a few pb for squarks and gluinos as heavy as 1 TeV. Their cascade decays would lead toa variety of signatures involving multi-jets, leptons, photons, heavy flavours and missing ener-gy. In several models, discussed in detail in this volume, the precision measurement of themasses of SUSY particles and the determination of the model parameters will be possible. Themain challenge would be therefore not to discover SUSY itself, but to reveal its nature and de-termine the underlying SUSY model.

Other searches beyond the Standard Model have been also investigated. Throughout this vol-ume are presented strategies for searching for technicolour signals, excited quarks, leptoquarks,new gauge bosons, right-handed neutrinos and monopoles. Given the large number of detailedmodels published in this field, the task of evaluating each of them is beyond the scope of thisdocument. Rather an exploratory point of view is taken, examples are used and in some cases adetailed study is performed.

14.4 Simulation of physics signals and backgrounds

In the process of evaluation of the physics potential of the ATLAS experiment, Monte Carloevent generators were used to simulate multiparticle production in physics processes appearingin the pp collisions. Detailed or parametrised simulation of the detector response to this multi-particle stream was then used to evaluate the possible observability of the signal.

In the full detector simulation, described in Section 2.2, the detailed geometry of the detector isimplemented and the interactions of particles with the material of the detector are modelled.Results from full-simulation studies have been described in Chapters 3-10 for several crucial

14 Physics overview 463

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

benchmark signatures and physics processes, e.g. mass resolutions, acceptances and identifica-tion efficiencies for H → γγ, H → ZZ∗ → 4l, H → bb, H → ττ decays, ETmiss resolution, b-jet and τ-jet identification capability.

However, in most of the cases presented in this volume, evaluation of the expected signals andbackgrounds has been done with the fast simulation described in Section 2.5. This simulationincludes, in a parametrised way, the main aspects related to the detector response: jet recon-struction in the calorimeters, momentum/energy smearing for leptons and photons, reconstruc-tion of missing transverse energy and charged particles. It is tuned to reproduce as well aspossible the expected ATLAS performance, and this tuning has been verified with severalbenchmark processes as described in Section 2.5.

The fast simulation was used very extensively for estimating the expected backgrounds fromphysics processes. Such approach was particularly useful for channels requiring large eventsamples, which one could not process with the much more time-consuming full simulation.Many of these studies are presented in this volume, e.g. for Higgs searches in Chapter 19, where,in some cases, both irreducible and reducible backgrounds required simulations of several mil-lion events.

14.4.1 Event generators

There are several available Monte Carlo event generators for pp collisions, the most exhaustiveones, with respect to available physics processes and complexity in modelling hadronic interac-tions, being: HERWIG [14-14], ISAJET[14-12] and PYTHIA[14-13]. Each of these simulates ahadronic final state corresponding to some particular model of the underlying physics. The de-tails of the implementation of the physics are different in each of these generators, however theunderlying philosophy of the generators is the same.

• The basic process is a parton interaction involving a quark or gluon from each of the in-coming protons. Elementary particles in the final state, such as quarks, gluons or W/Z/γ-bosons, emerge from the interaction. The fundamental process is calculated in perturba-tive QCD, and the initial momentum of the quarks or gluons is given by structure func-tions.

• Additional QCD (gluon) radiation takes place from the quarks and gluons that partici-pate in the basic scattering process. These parton showers are based on the expansionsaround the soft and collinear limits and can be ascribed to either the initial or final state.The algorithm used by HERWIG includes some effects due to quantum interference andgenerally produces better agreement with the data when detailed jet properties are stud-ied. The showering continues down to some low energy cut-off. For some particular casesthe matrix element calculations involving higher-order QCD processes are used. Theevents that have more energy in the parton process have more showering, and conse-quently more jet activity.

• The collection of quarks and gluons must then be hadronised into mesons and baryons.This is done differently in each of the event generators, but is described by a set of (frag-mentation) parameters that must be adjusted to agree with experimental results. HER-WIG looks for colour singlet collections of quarks and gluons with low invariant massand groups them together; this set then turns into hadrons. PYTHIA splits gluons intoquark-antiquark pairs and turns the resulting set of colour singlet quark-antiquark pairsinto hadrons via a string model. ISAJET simply fragments each quark independently pay-ing no attention to the colour flow. In ISAJET the underlying event that arises from the re-

464 14 Physics overview

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

maining beam fragments must be added. The other generators tie these fragments backinto the partonic system in order to neutralise the colour.

Matrix elements are likely to provide a better description of the main character of the events, i.e.the topology of well separated jets, while parton showers should be better at describing the in-ternal structure of these jets.

The above model(s) describe events where there is a hard-scattering of the incoming partons; ei-ther a heavy particle is produced or the outgoing partons have large transverse momentum.While these are the processes that are of most interest, the dominant cross-section at the LHCconsists of events with no hard scattering. There is little detailed theoretical understanding ofthese minimum-bias events and the event generators must rely on data at current energies.These minimum-bias events are important at LHC, particularly at design luminosity, as theyoverlap interesting hard-scattering events such as the production of new particles. The genera-tors use a different approach in this case. ISAJET uses a pomeron model that has some theoreti-cal basis. HERWIG uses a parametrisation of data mainly from the CERN pp Collider. PYTHIAuses a mini-jet model where the jet cross-section is used at very low transverse momenta, i.e thehard scattering process is extrapolated until it saturates the total cross-section. Whenever rele-vant, ATLAS has used the PYTHIA approach with dedicated modifications that agree withpresent data from Tevatron [14-17]. The multiplicity in minimum-bias events predicted by thisapproach is larger than that predicted by ISAJET or HERWIG (see Chapter 15), hence issues as-sociated with pile-up are treated conservatively.

The generators differ in the extent to which non-standard physics processes are included. Themost complete implementation of the Standard Model processes are available in PYTHIA, whileISAJET has the most complete implementation of SUSY scenarios.

In the physics evaluation presented in this volume, the Standard Model physics and Higgssearches were mostly simulated with PYTHIA. ISAJET was used extensively for the supersym-metry studies but some analyses have been done also with the supersymmetric extension of PY-THIA [14-23]. HERWIG has been used for some of the QCD studies. The model of the hadronicinteractions implemented in the physics generator has a direct impact on physical observablessuch as jet multiplicity, their average transverse momentum, internal structure of the jets andtheir heavy flavour content. That was one of the reasons why, whenever possible, PYTHIA wasused enabling a consistent set of signal and background simulations to be generated.

Theoretical precision of the existing Monte Carlo generators is far from adequate for the chal-lenging requirements of the LHC experiments. Despite the huge efforts which have been putinto developing of physics generators for hadron colliders over the last years, the precision withwhich e.g. present data can be reproduced is not better than 10-30%, and in some cases is notbetter than a factor of two.

Table 14-1 shows a few examples of important signal and background processes with their pre-dicted cross-sections, as used in the simulations discussed in this volume. If not explicitly statedotherwise, these are calculated using leading-order QCD as implemented in PYTHIA 5.7, usingthe CTEQ2L set of structure functions as the reference one. Whenever better or more appropri-ate calculations were available, the production cross-section from PYTHIA was suitably nor-malised, or a different Monte Carlo generator was used.

• The QCD multi-jet production is a dominant background for e.g. Higgs searches in themulti-jet final state. The production of events with three or more high-pT jets is not wellmodelled by lowest-order di-jet processes convoluted with parton showers. To illustratethe large discrepancy between exact matrix element calculations and parton shower ap-

14 Physics overview 465

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

proaches in this case, Table 14-1 gives rates for one, three and four jet final states as givenby the exact multi-parton matrix element NJETS Monte Carlo [14-15] and PYTHIA. Onthe other hand, heavy flavour content of jets is not modelled with the NJETS Monte Carlo.Simulation of four b-jet final states has been therefore only possible with the PYTHIA gen-erator, which has the heavy flavour content of the partonic shower implemented.

• In the case of di-jet production in association with a W or Z, the VECBOS Monte Carlo[14-16], dedicated to this process, has been used. Exact matrix-element calculations wereused also for estimating the expected cross-section in the case of Wbb [14-18] and Zbb [14-19] production. In the first case a modified version of HERWIG [14-18] was used, while inthe second case the EUROJET Monte Carlo [14-21] was adopted.

• The leading order tt cross-section is quoted in Table 14-1 since it has been used for all thebackground studies to new physics. For the specific case of top physics studies inChapter 18, a more accurate NLO calculation of 833 pb has been used, except for the caseof single-top production, for which the NLO terms are not yet known.

• The total bb cross-section is also quoted in Table 14-1. For the B-physics studies, muchmore detailed work reported in Chapter 17 has shown that for high-pT b-quark produc-tion, which can provide a Level-1 trigger with a high-pT muon, the PYTHIA model asused by ATLAS [14-20] reproduces quite well the bb production as measured at the Teva-tron [14-21] [14-22]. In this case, only a small fraction of the total cross-section quoted inTable 14-1 is relevant for physics, and many of the large theoretical uncertainties inherentto the calculations of the total bb production are very significantly reduced. A more de-tailed discussion of bb production at the LHC is discussed in Section 15.8.

The list above collects some relevant examples of the attempts which have been made to esti-mate as correctly as possible the expected production rates at the LHC. More details can befound in the specific Chapters of this volume discussing particular physics processes

Large uncertainties in the signal and background production cross-sections, due to missinghigher-order corrections, structure function parametrisations, energy scale for the QCD evolu-tion, as well as models used for full event generation, remain. In addition, despite the existenceof many higher-order QCD correction (K-factor) calculations, not all processes of interest at theLHC have benefited from this theoretical effort. In most cases they have also not been embodiedin the Monte Carlo generator, so that proper studies of their impact on the observed rates can-not be undertaken. Therefore, the present studies consistently and conservatively avoided theuse of K-factors, resorting to Born-level predictions for both signal and backgrounds.

14.4.2 Signal observability

In the following sections, most of the results will be given for integrated luminosities of 30 fb-1and 100 fb-1, which are expected to be collected in three years of data taking at the initial (low)luminosity and one year of data taking at the design (high) luminosity respectively. The ulti-mate discovery potential is evaluated for an integrated luminosity of 300 fb-1.

In most cases, the event selection for signal and background has been performed as it might beexpected for off-line analysis. The foreseen trigger LVL1/LVL2 menus were used for the dis-cussed channels. The possible irreducible and reducible backgrounds are extensively discussed.Given that the presently available tools for physics modelling have inherent uncertainties, anal-yses in most cases are straightforward; sophisticated statistical methods and very detailed opti-misation of cuts are not applied.

466 14 Physics overview

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

The observation of a given signal will be considered as possible if a significance of five standarddeviations, defined according to the naive estimator S/ , where S (B) is the expected numberof signal (background) events, can be obtained. This includes the relevant systematic uncertain-ties. If the number of expected signal and background events is smaller than 25, Poisson statis-tics has been used to compute the equivalent Gaussian significance.

Table 14-1 Leading order cross-sections for some typical processes at the LHC. Unless stated otherwise, thesenumbers have been obtained by using PYTHIA 5.7 with CTEQ2L structure functions.

Process Cross-section Comments

Inclusive HmH = 100 GeV

27.8 pb

WH with W → lνmH = 100 GeV

0.40 pb

ttH with one W → lνmH = 100 GeV

0.39 pb

Inclusive SUSY, ~ 1 TeV

3.4 pb ISAJET or PYTHIA

Inclusive bb 500 µb All di-jet processes used

Inclusive tt (mt = 175 GeV) 590 pb

Di-jet processes:1 jet pTj > 180 GeV, |η| < 3.2

3 jets pTj > 40 GeV, |η| < 3.2

4 jets pTj > 40 GeV, |η| < 3.2

13 µb

2.0 µb (0.7 µb)

0.4 µb (0.1 µb)

PYTHIA

NJETS (PYTHIA)

NJETS (PYTHIA)

Inclusive W 140 nb

Inclusive Z 43 nb

Wjj with W → lνwith 2 jets pTj > 15 GeV, |η|< 3.2

4640 pb VECBOS

Wbb with W → lν 69.3 pb Matrix element [14-18]+ HERWIG

Zjj with Z → llwith 2 jets pTj > 15 GeV, |η|< 3.2

220 pb VECBOS

Zbb with Z → ll 36 pb EUROJET + [14-19]

WW 71 pb

WZ 26 pb

Wγ with W → lνwith pTγ > 100 GeV, |η| < 2.5

210 fb

mg̃ mq̃

B

14 Physics overview 467

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

14.5 Outline

This volume reviews the potential of the ATLAS detector for the observability of a variety ofphysics processes, starting from the studies of hadronic physics, precision measurements in theStandard Model sector and CP-violation phenomena, continuing through the searches for theHiggs boson(s) and supersymmetry, and ending with a discussion of physics beyond the Stand-ard Model.

The volume begins with a discussion of QCD processes (Chapter 15), which have the largestrate and represent the dominant background for new physics searches. Next is a discussion ofthe properties of the W and Z gauge bosons and how ATLAS can improve the precision meas-urements of the masses and couplings (Chapter 16). This is followed by a presentation of the B-physics programme; methods for the measurement of CP violation, mixing and rare decays arediscussed (Chapter 17). Next, measurements related to the top quark and searches for otherheavy quarks/leptons are described (Chapter 18). The Standard Model Higgs boson and its var-iants in the minimal supersymmetric model provide a benchmark for LHC physics; the largenumber of possible discovery channels are analysed in detail (Chapter 19). Physics beyond theStandard Model is the subject of the final two sections; the most popular extension to Super-symmetry is discussed in detail and many signatures that allow precise measurements in thissector are presented (Chapter 20). Finally, signatures for other extensions to the Standard Mod-el, such as new gauge bosons and technicolour, are discussed (Chapter 21).

14.6 References

14-1 ATLAS Letter of Intent, CERN/LHCC/92-4, CERN 1992.

14-2 ATLAS Technical Proposal, CERN/LHCC 94-43, CERN 1994.

14-3 S. Glashow, Nucl. Phys. 22 (1961) 579;S. Weinberg, Phys. Rev. Lett. 19 (1967) 1264;A. Salam, in: ‘Elementary Particle Theory’, W. Svartholm,ed., Almquist and Wiksell,Stockholm,1968;H.D. Politzer, Phys. Rev. Lett 30 (1973) 1346;D.J. Gross and F.E. Waltzed, Phys. Rev. Lett. 30 (1973)1343.

14-4 H. Fritzsh and M. Gell-Mann, Proc. XVI Int. Conf. on High Energy Physics, eds. J. D.Jackson and A. Roberts (Fermilab 1972).

14-5 M. Kobayashi and T. Maskawa, Prog. Theor. Phys. 49 (1973) 652;N. Cabibbo, Phys. Rev. Lett. 10 (1963) 531.

14-6 P. W. Higgs, Phys. Rev. Lett. 12 (1964) 132; Phys. Rev. 145 (1966) 1156;F. Englert and R. Brout, Phys. Rev. Lett 13 (1964) 321;G. S. Guralnik, C. R. Hagen and T. W. Kibble, Phys. Rev. Lett 13 (1964) 585.

14-7 L. Maiani, G. Parisi and R. Petronzio, Nucl. Phys. B136 (1979) 115;N. Cabbibo, L. Maiani, G. Parisi and R. Petronzio, Nucl. Phys. B158 (1979) 295;R. Dashen and H. Neuberger, Phys. Rev. Lett. 50 (1983) 1897;D. J. E. Callaway, Nucl. Phys. B233 (1984) 189;M. A. Beg, C. Panagiatakopolus and A. Sirlin, Phys. Rev. Lett 52 (1984) 883;M. Lindner, Z. Phys. C31 (1986) 295.

14-8 T. Hambye and K. Riesselmann, Phys. Rev. D55 (1997) 7255.

468 14 Physics overview

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

14-9 C. Quigg, B.W. Lee and H. Thacker, Phys. Rev. D16 (1977) 1519;M. Veltman, Acta Phys. Polon. B8 (1977) 475.

14-10 J. Wess and B. Zumino, Nucl. Phys. B70 (1974) 39.

14-11 For a review, see K.D. Lane hep-9605257 (1996).

14-12 F. Paige and S. Protopopescu, in Supercollider Physics, p. 41, ed. D. Soper (WorldScientific, 1986).

14-13 T. Sjostrand, Comp. Phys. Comm. 82 (1994) 74.

14-14 G. Marchesini et al., Comp. Phys. Comm. 67 (1992) 465.

14-15 F.A. Berends and H. Kuijf, Nucl.Phys.B353 (1991) 59.

14-16 F.A. Berends, H. Kuijf, B. Tausk and W.T. Giele, Nucl.Phys.B357 (1991) 32;W. Giele, E. Glover, D. Kosower, Nucl. Phys. B403 (1993) 633.

14-17 The CDF Collaboration, F. Abe et al., Phys. Rev. D41 (1990) 2330; Phys. Rev. Lett. 61 (1988)1819.

14-18 M.L. Mangano, Nucl. Phys. B405 (1993) 536.

14-19 B. van Eijk and R. Kleiss, in [14-24], page 183.

14-20 S. Baranov and M. Smizanska, ‘Beauty production overwiew from Tevatron to LHC’,ATLAS Internal Note ATL-PHYS-98-133 (1998);P. Eerola,’ The inclusive muon cross-section in ATLAS’, ATLAS Internal Note ATL-PHYS-98-120 (1998).

14-21 D0 Collaboration, S. Abachi et al., Phys. Rev. Lett. 74 (1996) 3548; Phys. Lett. B370 (1996)239.

14-22 CDF Collaboration, F. Abe et al., Phys. Rev. Lett. 71 (1993) 500; Phys. Rev. Lett. 71 (1993)2396; Phys. Rev. Lett. 71 (1993) 2537; Phys. Rev. Lett. 75 (1995) 1451; Phys. Rev. D50 (1996)4252; Phys. Rev. D53 (1996) 1051.

14-23 T. Sjostrand, Comp. Phys. Comm. 82 (1994) 74. The supersymmetry extensions aredescribed in S. Mrenna, Comp. Phys. Comm. 101 (1997) 232.

14-24 Proceedings of the Large Hadron Collider Workshop, Aachen, 1990, edited by G. Jarlskogand D. Rein, CERN 90-10/ECFA 90-133.

14 Physics overview 469

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

470 14 Physics overview

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

15 QCD processes at the LHC

15.1 Introduction

The study of QCD processes at the LHC will serve two main goals. First the predictions of QCDwill be tested and precision measurements will be performed, allowing additional constraints tobe established e.g. on the distribution of partons in the proton, or providing measurements ofthe strong coupling constant αs at various scales. Second QCD processes represent a major partof the background to other Standard Model processes and signals of new physics at the LHCand thus need to be understood precisely in the new kinematic region available here. Devia-tions from the QCD expectations might themselves also indicate the occurrence of new physics,as in the case of compositeness for the jet transverse energy and di-jet invariant mass and angu-lar distributions. Furthermore, the production cross-sections for almost all processes are con-trolled by QCD.

Tests of QCD can be performed by comparing measurements to fixed order (either LO (leadingorder) or NLO (next-to-leading order)) calculations or to leading-log Monte Carlo programswhich contain LO matrix elements and approximate higher orders through the use ofparton showers (and also include the hadronisation of the partonic system). Perturbative QCDcan also be tested by extracting (or constraining) the fundamental parameter αs. The differencebetween a LO and a NLO calculation is quantified in the K-factor; the K-factor is defined as theratio between the cross-section at NLO to the one at LO. The K-factor can become significantlylarger than 1, especially when new sub-processes appear at next-to-leading order. Calculationsat next-to-leading order are mostly restricted to parton level and often performed by numericalintegration of the corresponding matrix elements.

This chapter gives an overview of different measurements of QCD processes [15-1], [15-2], [15-3] to be performed with ATLAS, classified by the main characteristics (or main selection criteria)of the final state. Besides a qualitative overview, a few examples are given where first quantita-tive investigations of the potential of ATLAS have been performed. The organisation of thechapter is as follows: the next section contains a brief summary on the present knowledge ofparton densities and some perspectives for improvements before the start of LHC. Then meas-urements of properties of minimum-bias events (Section 15.3) are discussed, followed by a de-scription of studies of hard diffractive scattering (Section 15.4). Next, the information to bededuced from the measurement of jets (Section 15.5) is described, followed by a section on pho-ton physics (Section 15.6) and one concerning the production of Drell-Yan pairs and heavygauge bosons (Section 15.7). Before concluding, the production of heavy flavours (charm, bot-tom and top, Section 15.8) is discussed.

Unless stated differently in the corresponding sections, the standard trigger settings have beenused. The signatures listed in [15-4] for the first level (and the second level) of the trigger systemconsist mainly of inclusive signatures. It is foreseen to accept a fraction of events with lowerthresholds and it is possible to include specific signatures (esp. at the higher levels of the triggersystem) combining different objects and thus allowing lowering of the corresponding thresh-olds. One important exception is the case of hard diffraction and the case of minimum-biasevents, where dedicated triggers will have to be employed.

2 2→

15 QCD processes at the LHC 471

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

15.2 Knowledge of the proton structure

15.2.1 Global parton analyses and parton kinematics at the LHC

The calculation of the production cross-section at the LHC both for interesting physics process-es and their backgrounds relies upon a knowledge of the distribution of the momentum fractionx of the partons in the proton in the relevant kinematic range. These parton distribution func-tions (pdf’s) are determined by global fits (see [15-5] for a pedagogical overview) to data fromdeep-inelastic scattering (DIS), Drell-Yan (DY), jet and direct photon production at current ener-gy ranges. Two major groups, CTEQ [15-6] and MRS [15-7], provide regular updates to the par-ton distributions when new data and/or theoretical developments become available.

Lepton-lepton, lepton-hadron and hadron-hadron interactions probe complementary aspects ofperturbative QCD (pQCD). Lepton-lepton processes provide clean measurements of αs(Q2) andof the fragmentation functions of partons into hadrons. Measurements of deep-inelastic scatter-ing structure functions (F2,F3) in lepton-hadron scattering and of lepton pair production cross-sections in hadron-hadron collisions provide the main source of information on quark distribu-tions qa(x,Q2) inside hadrons. At leading order, the gluon distribution function g(x,Q2) enters di-rectly in hadron-hadron scattering processes with direct photon production and jet final states.Modern global parton distribution fits are carried out to next-to-leading order (NLO) which al-lows qa(x,Q2), g(x,Q2) and the strong coupling αs(Q2) to all mix and contribute in the theoreticalformulae for all processes. Nevertheless, the broad picture described above still holds to somedegree in global pdf analyses. In pQCD, the gluon distribution is always accompanied by a fac-tor of αs, in both hard scattering cross-sections and in the evolution equations for the parton dis-tributions. Thus, the determination of αs and the gluon distribution is, in general, a stronglycoupled problem. One can determine αs separately from e+e- interactions or determine αs andg(x,Q2) jointly in a global pdf analysis. In the latter case, though, the coupling of αs and thegluon distribution may not lead to a unique solution for either (see e.g. in [15-8]).

Currently, the world average of αs(MZ) is of the order of 0.118 − 0.119 [15-9]. The average valuefrom LEP is 0.121 while the DIS experiments prefer a somewhat smaller value (of the order of0.116 − 0.117). Since global pdf analyses are dominated by the high statistics DIS data, theywould favour the values of αs closer to the lower DIS values. The more logical approach is toadopt the world average and concentrate on the determination of the pdf’s. This is what bothCTEQ and MRS currently do. One can either quote a value of αs(MZ) or the value of ΛQCD. Forthe latter case, however, the renormalisation scheme used together with the number of flavourshas to be clearly specified. Usually the MS scheme is used. The specification of the number offlavours is important as the value of αs has to be continuous across flavour thresholds. A rangeof αs(MZ) of 0.105 to 0.122 corresponds to the range of 100 < ΛQCD < 280 MeV for five flavoursand to 155 < ΛQCD < 395 MeV for four flavours.

The data from DIS, DY, direct photon and jet processes utilised in pdf fits cover a wide range inx and Q. The kinematic ‘map’ in the (1/x,Q) plane of the data points used in a recent parton dis-tribution function analysis is shown in Figure 15-1. The HERA data (H1 and ZEUS) are predom-inantly at low x, while the fixed target DIS and DY data are at higher x. There is considerableoverlap, however, with the degree of overlap increasing with time as the statistics of the HERAexperiments increases. The DGLAP equations [15-10] in pQCD describe the change of the par-ton distributions with Q2. The NLO DGLAP equations should describe the data over the wholekinematic range shown in Figure 15-1. At very low x, however, the DGLAP evolution is be-lieved to be no longer applicable and a BFKL [15-11] description must be used. No clear evi-

472 15 QCD processes at the LHC

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

dence of BFKL physics is seen in the current range of data; thus all global analyses useconventional DGLAP evolution of the pdf’s. There is a remarkable consistency between thedata in the pdf fits and the NLO QCD theory to fit these. Over 1300 data points are shown inFigure 15-1 and the χ2/DOF for the fit of theory to data is of the order of 1.

In Figure 15-2 the kinematics appropriate forthe production of a state with mass M and ra-pidity y at the LHC is shown [15-12]. For ex-ample, to produce a state of mass 100 GeV atrapidity y = 2 requires partons of x values 0.05and 0.001 at a Q2 value of 104 GeV2. The figurealso shows another view of the kinematic cov-erage of the fixed target and the HERA experi-ments used in the pdf fits.

15.2.2 Properties and uncertainties ofparton distribution functions

Figure 15-3 shows the parton distributions forthe different quark flavours and the gluon asobtained from the CTEQ4M distribution [15-8]for a scale of Q2 = 20 GeV2, in Figure 15-4 thecorresponding distributions are shown for ascale of Q2 = 104 GeV2. Clearly visible is thedominance of the gluon distribution for smallparton momenta. In addition the violation ofthe flavour symmetry for u and d sea quarkscan be seen.

Figure 15-1 A kinematic map of data points in the (1/x,Q) plane from different processes used in a global fit ofparton densities (from [15-5]).

100 101 102 103 104

1/X

100

101

102

Q (

GeV

)DIS (fixed target)HERA (’94)DYW-asymmetryDirect-γJets

Figure 15-2 Parton kinematics at the LHC (from [15-12]) in the (x,Q2) kinematic plane for the production ofa particle of mass M at rapidity y (dotted lines).

10-7 10-6 10-5 10-4 10-3 10-2 10-1 100100

101

102

103

104

105

106

107

108

109

fixedtarget

HERA

x1,2 = (M/14 TeV) exp(±y)

Q = M

M = 10 GeV

M = 100 GeV

M = 1 TeV

M = 10 TeV

66y = 40 224

Q2

(G

eV2 )

x

15 QCD processes at the LHC 473

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

Parton distribution determined at a given x and Q2 ‘feed-down’ to lower values of x at highervalues of Q2. The accuracy of the extrapolation to higher Q2 depends both on the accuracy ofthe original measurement and any uncertainty on αs(Q2). For the structure function F2, the typi-cal measurement uncertainty at medium to large x is of the order of 3%. At high Q2 (about105 GeV2) there is an extrapolation uncertainty of 5% in F2 due to the uncertainty in αs.

Figure 15-6 shows the gluon distribution as afunction of x for five different values of Q2, us-ing the CTEQ4M distribution. Most of the evo-lution takes place at low Q2 and there is onlylittle evolution for x values around 0.1. In con-trast, at an x value of 0.5, the gluon distribu-tion decreases by a factor of approximately 30from the lowest to the highest Q2.

Global fits can also be performed using lead-ing-order (LO) matrix elements, resulting inleading-order parton distribution functions.Such pdf’s are preferred when leading ordermatrix element calculations (such as in MonteCarlo programs like HERWIG [15-13] and PY-THIA [15-14]) are used. The differences be-tween LO and NLO pdf’s, though, areformally NLO; thus the additional error intro-duced by using a NLO pdf should not be sig-nificant. A comparison of the LO and NLOgluon distribution is shown in Figure 15-7 forthe CTEQ4 set, where the LO distribution is CTEQ4L and the NLO distribution is CTEQ4M.The differences get even smaller at larger Q2 values.

Figure 15-3 Parton distributions for the CTEQ4M pdfat Q2 = 20 GeV2. The gluon distribution has beenreduced by a factor of 10.

Figure 15-4 Parton distributions for the CTEQ4M pdfat Q2 = 104 GeV2. The gluon distribution has beenreduced by a factor of 10.

0

0.5

1

1.5

2

10-4

10-3

10-2

10-1

1x

xf(x

,Q2 )

0

1

2

3

4

10-4

10-3

10-2

10-1

1x

xf(x

,Q2 )

Figure 15-5 Gluon distribution for the CTEQ1L,CTEQ2L, CTEQ3L and CTEQ4L pdf’s at a value ofQ2 = 5 GeV2 (from [15-5]).

10-4

10-3

10-2

10-1

1

10

10-3

10-2

10-1

x

x*G

(x,Q

2 )

Parton Density of the Nucleon

Gluon Distribution

Q2 = 5 GeV2

CTEQ4LCTEQ3LCTEQ2LCTEQ1L

474 15 QCD processes at the LHC

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

Many of the comparisons in this document have been performed with the CTEQ2L pdf, a pdfthat is on the order of five years old [15-15]. A comparison of the gluon distribution forCTEQ1L, CTEQ2L, CTEQ3L and CTEQ4L is shown in Figure 15-5. With increasing amounts ofdata included from HERA, the tendency has been for the low x pdf’s to increase. The relative in-creases are reduced at higher values of Q2.

Figure 15-6 Gluon densities as a function of x fromthe CTEQ4M parton distribution set for five differentQ2 values: 2, 10, 50, 104 and 106 GeV2 (from [15-5]).

Figure 15-7 Comparison of the gluon distributionfrom the CTEQ4L (leading order) and the CTEQ4M(next-to-leading order) global fit (from [15-5]).

Figure 15-8 Normalised quark-gluon luminosity func-tion (as a function of ) for variations inthe gluon distribution which are consistent with exist-ing DIS and DY datasets (from [15-17]). The dottedcurve shows a toy model with more quarks at x > 0.5for large Q2 than in CTEQ4M.

Figure 15-9 Normalised gluon-gluon luminosity func-tion (as a function of ) for variations inthe gluon distribution, which are consistent with exist-ing DIS and DY datasets (from [15-17]).

10-4

10-3

10-2

10-1

1

10

10 2

10-3

10-2

10-1

x

x*G

(x,Q

2 )

Parton Density of the Nucleon

CTEQ4M Gluon Distribution

Q2 = 2 GeV2

Q2 = 10 GeV2

Q2 = 50 GeV2

Q2 = 104 GeV2

Q2 = 106 GeV2

10-4

10-3

10-2

10-1

1

10

10-3

10-2

10-1

x

x*G

(x,Q

2 )

Parton Density of the Nucleon

Gluon Distribution

Q2 = 5 GeV2

CTEQ4MCTEQ4L

τ x1 x2⋅= τ x1 x2⋅=

15 QCD processes at the LHC 475

ATLAS detector and physics performance Volume IITechnical Design Report 25 May 1999

In addition to having the best estimate for the values of the pdf’s in a given kinematic range. itis also important to understand the allowed range of variation in the pdf’s, i.e. their uncertain-ties. The conventional method of estimating parton distribution uncertainties is to compare dif-ferent published parton distributions. This is unreliable since most published sets of partondistributions (e.g. from CTEQ and MRS) adopt similar assumptions and the differences betweenthe sets do not fully explore the uncertainties that actually exist. Ideally, one might hope to per-form a full error analysis and provide an error correlation matrix for all the parton distributions(see e.g. [15-16]). This goal may be difficult to carry out for two reasons. Experimentally, only asubset of the experiments usually involved in the global analyses provide correlation informa-tion on their data sets in a way suitable for the analysis. Even more important, there is no estab-lished way of quantifying the theoretical uncertainties for the diverse physical processes thatare used and uncertainties due to specific choices of parametrisations. Both of these are highlycorrelated.

As the LHC is essentially a gluon-gluon collider and many hadron collider signatures of physicsboth within and beyond the Standard Model involve gluons in the initial state, it is important toestimate the theoretical uncertainty due to the uncertainty in the gluon distribution. The mo-mentum fraction carried by gluons is 42% with an accuracy of about 2% (at Q = 1.6 GeV in theCTEQ4 analysis), determined from the quark momentum fraction using DIS data. This impor-tant constraint implies that if the gluon distribution increases in a certain x range, momentumconservation forces it to decrease in another x range. To estimate the uncertainty on the gluondistribution, an alternative approach has been carried out [15-17]: the (four) parameters of thegluon distribution (based on the CTEQ4 set) have been varied systematically in a global analy-sis and the resulting parton distributions have been compared to the DIS and Drell-Yan datasetsmaking up the global analysis database. Only DIS and Drell-Yan datasets were used, as the ex-perimental and theoretical uncertainties for these processes are under good control. Only thosepdf’s that do not clearly contradict any of the (DIS and Drell-Yan) data sets in the global analy-sis database were kept. The variation of the gluon distribution obtained with this proced