SMandjetmeasurementsattheLHC
AndyPilkington–IPPPandManchesterPresentedattheYoungExperimentandTheoryIns7tute,Durham,January2011
PartI–jetcharacteris.cs1)Introduc?ontoATLASandtheLHC2)Jetreconstruc?onandperformance3)Inclusivejetmeasurements
PartII–eventstructure4)Mul?‐jetfinalstates5)Vectorbosonproduc?on(with/withoutjets)6)Whattoexpectfrom2011
TheLHCin2010
• 45pb‐1ofdatatakenintotal.• Averagenumberofinterac?onsperbunchcrossingwasabout3.5forthefinalruns.
• Theplanistotakeatleast1U‐1ofdatain2011at√s=7TeV.– Likelythattheluminositywillbehigherthanthis,possiblyevenincreaseinenergy
to8‐9TeV(underdiscussion)
TheATLASdetector–ageneralmul?purposedetector
JetsatATLAS:reconstruc?onandperformancein2010
TheATLAScalorimeter
Thean?‐kTalgorithm
• DefaultalgorithmforstudyingjetsatbothATLASandCMS.
• Foreachobject,‘i’,andeachpairofobjects,‘ij’,definethemergingparameters:
where
• Ifdijissmallest,mergeobjects‘i’and‘j’• IfdiBissmallest,defineobject‘i’asajetand
removefromthelist.• Repeatprocedureun?lnoobjectsremaining.
• Algorithmisbothinfra‐redandcollinearsafe.• Generallyproducesjetsofdefinitesize
JHEP04(2008)063
Jetinputs
• Topo‐clustersare3Dobjects:
– Clusteringacrossmanylayersofcalorimeter– Seedcelliden?fiedby|E|>4σini?atesthe
cluster– Neighbourcellsitera?velyaddedtoclusterif
|E|>2σ– RingofGuardcellsadded(|E|>0σ)
• Towersareconstructedbyusingcellsthatarealignedalongagivenη‐ϕdirec?on– 6400towersinATLAS,– Noisesuppressedtowersarethoseconstructed
fromcellsthatbelongtotopo‐clusters
Jetcalibra?onschemes
Defaultjetenergyscalecalibra?onanduncertainty
• Thedefaultcalibra?onschemecurrentlycorrectsEM‐scalejets.
• Response,R,ofajetinthecalorimeterisstudiedforjetsinPYTHIA:
• Responseparameterizedandinvertedtogivethejetenergyscale(JES)correc?on
• UsingtheMCderivedresponseintroducesuncertaintydueto:
– Detectorsimula?on(e.g.geometry,deadmaterial,noise,hadronicshowermodels…)
– Eventgeneratormodeling(hadroniza?on,partonshowerandunderlyingevent)
pEMT = R(ptruth
T , η) ptruthT
Theproblemwithforwardjets
• Therela?veresponseasafunc?onofpseudo‐rapidityismeasuredin‐situ:
– Dijeteventsselected– ISR/FSRsuppressedbyselec?ngback‐to
backjets(|Δϕ|>2.6)andcuongonthirdjetpT
– Asymmetryofdijetsmeasuredtogivetherela?veresponse
• DataandMCdonotagreewellintheforwardregion!Couldbe:
– Detectorsimula?onproblem(shouldcalibrate).
– Aphysicsmodelingproblem(shouldnotcalibrate
Thedifferencebetween‘quark’and‘gluon’jets
• Jetsstudiedsofarincludejetsorigina?ngfrombothhardgluonsandhardquarksintheMC:– jetsfromgluonstypicallysoserandsospar?clestypicallydepositlessenergy
• Thiscausesaproblemifthesampleofjetswearelookingatdonothavethesamemixtureofquarks/gluonsastheinclusiveMCsample– Addi?onaluncertaintyneededthatisanalysisspecific(no‘onecalibra?onfitsall’approach)
Theproblemofnear‐byjets
• Thejetcalibra?onwascalculatedforjetsthatarewellseparatedfromanyotherjetac?vity
• Responseismuchlowerinjetsifthereare‘nearby’jets• Addi?onaluncertaintyinmul?‐jetanalyses
– dependsonthetopologyoftheeventinques?on.
Howgoodisthiscalibra?on?–currentATLASview
• Valida?onoftheenergyscalewasperformedwiththesinglehadronresponse,measuringtheE/pforchargedpar?cles.Liuledevia?onbetweendataandMC.
• Propaga?ngthisuncertaintytothejet‐level[usingpseudo‐experiments],yieldsamuchsmallerintrinsicuncertaintythanthedefaultMC‐basedapproach.
Howgoodisthiscalibra?on?–somepre‐dataexpecta?ons
• Themeasurementoftheproduc?oncross‐sec?onofVBFHiggsiscrucialtotheextrac?onoftheHiggscouplings.DominantuncertaintyfromJES.
– TheCSCstudyassumed7%(15%)JESuncertaintyforcentral(forward)jets.– Wealreadyhavethatprecision!
• Themeasurementofthetop‐quarkmassinsemi‐leptonicandhadronicchannelsisdominatedbyJES
– 0.7GeVuncertaintyperpercentJESuncertainty.– Alongwaytogofortop‐studies.
Howgoodisthiscalibra?on?–comparisonwithCMS
• TheCMSbasicjetenergyscaleissmallerthanthecurrentATLASoneinthecentralregion– Thismakessense,astheyhaveused2.9pb‐1ofdataandphoton‐jetbalancingtogetan
in‐situhandleontheJES.TheATLASJESwillbesimilar.• Intheforwardregion,theCMSuncertaintyisMUCHsmallerthantheATLASone.
– MC/datadiscrepancyiscalibratedawayaserextrapola?ngthediscrepancytopT,3=0– ThisassumesthatthediscrepancyisNOTduetophysicseffects.ATLASdoesnotauempt
toclaimthisyet.
Howgoodisthiscalibra?on?–comparisonwithD0
• Itshouldn’tcomeasasurprisethattheJESatD0isbeuerthanthatatATLASandCMS
• Centraluncertaintydeterminedbyphoton‐jetbalancetobe<2%forjetsinthecentralregion.
• Uncertaintygrowstoabout4%by|η|~3
• TheATLAS/CMSgoalofa1%uncertaintyintheJESisrealis?c(atleastforcentraljets)
InclusivejetmeasurementsatATLAS
Inclusivejetmeasurements–thebenefits
• Keyobservableinpar?clephysics– previouslymeasuredine+e‐,ep,ppandcolliders
• Theinclusivemeasurementisanidealsteppingstonetomorecomplicatedjettopologies.– Simplestpossiblejetmeasurement
– NLOpredic?onknownforover20years,withsmalluncertainty
• Oncetheinclusivejetmeasurementhasbeenmadetohighaccuracy,canuseto– provideprecisemeasurementofαs
– determinethestructureoftheproton(PDFfits)– studyBSMphysics(seetalkbyC.Issever)
pp̄
Eventselec?on
• An?‐kTalgorithmusedwithR=0.4andR=0.6andcomparedtoQCDcalcula?ons
– Differentsosphysicseffectsfordifferentjetradii
• Inclusivecross‐sec?onmeasuredasafunc?onofjetpTandyforalljetswithpT>60GeVand|y|<2.8
– Alljetsinregionofcentraljet‐triggerplateau– Noforwardjets(MC/dataJESdiscrepancy)
• Dijetcross‐sec?onmeasuredasafunc?onofmjjandχ=exp(|y1‐y2|)– Leadingjetsa?sfiespT>60GeVand|y|<2.8– Secondjetsa?sfiespT>30GeVand|y|<2.8– Highpurityandefficiencyofjetreconstruc?on
forbothjets– StabilityofNLOimprovedbyasymmetriccuts
Also:1) Vertexrequiredwith5+tracks2) Jetscleanedtoremovejetsfrom
‘noise’andpoorlyreconstructedregionsofdetector
Unfoldingdetectoreffects
• Bin‐by‐binunfoldingtoremoveeffectofdetectorinefficienciesandresolu?ons.• Unfoldingfactor(σtruth/σreco)determinedusingPYTHIA6,uncertaintyonthisevaluated
by(i)usingdifferentgenerators(ii)changingtheresolu?on(iii)changingtheshapeoftheinputdistribu?ons.
• Typically5%(orless)uncertaintyinunfoldingfactors.
Inclusivejets:comparisontoLOeventgenerators
• ShapeiswelldescribedbytheLO2‐>2matrixelement+partonshowerapproach• Cross‐sec?onsdeterminedatleadingorder–normaliza?onrequiredtogettheMCto
matchthedata
Inclusivejets:comparisonwithNLOpredic?ons
• NLOpredic?onatparton‐levelfromNLOJet++usingCTEQ6.6PDF(checkedwithJETRAD)• Sos(non‐perturba?ve)correc?onsappliedtopartonlevelpredic?onstoallowcomparison
todata
ComparisonwithNLOpredic?ons,inabitmoredetail
Approximately5%uncertaintyinNLOtheory,comparedto30%experimentaluncertainty‐thisisJESdominated,shouldbeabletodras?callyimproveonthisinthenearfuture
Dijetmassdistribu?ons
Dijetangulardistribu?ons
Inclusivejetcross‐sec?oncomparisonwithPOWHEG
• POWHEG‐Boxrecentlyimplementeddijetproduc?on(publishedinarXiv:1012.3380)• NLOmatrixelementinterfacedtoPYTHIA/HERWIGtobenefitfrompartonshower,
hadroniza?onandunderlyingevent.– Allowsout‐of‐boxpar?clelevelcomparison(shownabove)
• But,fortheexperiments,POWHEGcouldbeusedto– improvetheuncertaintyintheunfoldingcorrec?ons
– ReducephysicsmodelingdependenceofJESdetermina?onusingdijetbalancing
Summary–partI
Themeasurementofjetsinacolliderexperiment1) Alotofworkisneededtocalibratetheenergyofjetsinacolliderexperimenttothe
accuracyrequiredforprecisionmeasurements2) ThecurrentpublicATLAS(CMS)uncertaintyisabout7%(4%)forcentraljetsandis
improvingsteadilythroughin‐situmeasurements.
Theinclusivejetcross‐sec.onmeasurement1) Thesimplestmeasurementofjetproduc?onattheLHC‐asteppingstonetothemore
complicated,exclusivefinalstatesdiscussedinthenextlecture2) DominatedbyJESuncertainty(>30%)3) NLOpredic?onsaccuratetoabout5%intheinclusivesinglejetmeasurement
Eventwithdijetmass,mjj=3.1TeV–theSMatunprecedentedenergies
Extras
90 91 92 93 94 95 96 97 98 99 100
MUON-CSC
MUON-DT
MUON-RPC
HCAL BARREL
HCAL ENDCAP
HCAL FORWARD
ECAL BARREL
ECAL END-CAP
PRE-SHOWER
STRIP TRACKER
PIXEL TRACKER
MUON-CSC
MUON-DT
MUON-RPC
HCAL BARREL
HCAL ENDCAP
HCAL FORWAR
D
ECAL BARREL
ECAL END-CAP
PRE-SHOWE
R
STRIP TRACKE
R
PIXEL TRACKE
R Series1 98.5 99.8 98.8 99.9 100 99.9 99.3 98.9 99.8 98.1 98.2
StatusoftheCMSdetectors
Thelimita?onsoftheinclusivejetmeasurements
• Lowsta?s?cs:measurementperformedwithlessthan1/1000thofthecurrentavailabledata.
• Conserva?veJESes?mate–resultsinexperimentaluncertainty6xlargerthanNLOuncertainty
• CoversalimitedpTrange(60<pT<500GeV)andlowrapidityrange(|y|<2.5).
• Futureextensions(probablyforthewinterconferences):– ImprovedJESonitsway!
– Fulldatasetof2010,allowinglargerpTandmjjtobeprobed– AswellasprobingsmallerpT(aslowas20GeV)andlarger|y|(upto4.5)