DecayoftheTau MajorLeptonic DecayChannels MajorHadronic - - PowerPoint PPT Presentation

DecayoftheTau

MajorLeptonic DecayChannels MajorHadronic DecayChannels

Onechargedtrackinjet=49.5% 17.84± 0.5% 17.36± 0.5%

mτ =1.78GeV/c2 cτ =87.m

Level1Tau TriggerusingtheHadronic andElectromagnetic Calorimeters1 I

1. Takea12x12groupingoftriggertowersaroundanactivespotasagenericjet. 2. Each4x4regionshowedhasaτ1vetobitassociatedwithit.Ifmorethan2triggertowersinthe ECALorHCALareactivated,thisτ1vetobitisset.Eachtriggertowerhasaprogrammablethreshold

• ftypicallyafewGeVs.

3. Agenericjetisrelabelled aTau Jetifnoneofitsnine4x4regionshaveaτ1vetobiton. 4. ThefourhighestET jetsandthefourhighestET jetsidentifiedasTau Jetsaresenttotheglobal triggerforconsiderationastriggerobjects.

Level1Tau TriggerusingtheHadronic andElectromagnetic Calorimeters1 II

TheefficiencyofTau1JetidentificationusingLevel1 algorithmisplottedagainstthetransverseenergyof thejet. ThedropwithhighETmaybebecausehigher energiescauseneighbouring triggertowerstofire

• ccassionally.

TheefficiencyofTau1JetidentificationusingLevel1 algorithmisplottedagainstpseudorapidity. Itfallsoffathighpseudorapidity presumablybecause theTau1Jetsareincidentonthecalorimetersathigher incidentanglesandtriggermultipletowers.

Level2Tau TriggerusingtheElectromagneticCalorimeter1 I

• Tau jetsaremorecollimatedthanQCDjets.

Define:<R2 =<η2 +<Φ2 Weexploitthiscollimationbyrequiringthatthe transverseenergydepositedintheneighbourhood (0.13<R<0.4)ofthehighesttowerbebelowacutoff energyPisol

cut.Thatis:

cut isol isol R T R T

P P E E < = − ∑

∑

<

• <
• 13

. 4 .

Efficiencyofacceptingsimulation1generatedTau and QCDjetsfordifferentET werestudiedusingadetector simulation.Pisol

cut wasvariedfrom0GeV/c to14GeV/c

andefficiencieswereplottedagainstitasshowntothe left. ItwasseenthatQCDjetsarebestrejectedforPisol

cut =5

Gev.

Level2Tau TriggerusingtheElectromagneticCalorimeter1 II

Asupermodule sectionoftheECALunderconstructionisshowntothe left.EachcrystalcrosssectioncorrespondstoR=9.2x1018,well abovetheRresolutionrequiredforthedefinitionofPisol

cut.

Pisol

cut =5GeV/c isusedinaMonteCarlotest.

AplotbetweentheTau jetidentificationefficiencyintheECALversus transverseenergyisshowntotheright.Theefficienciesofthe different channelsofhadronic decayareseparated. AplotbetweentheTau jet identificationefficiencyversusη isshowntotheright.Thedropin efficiencywithincreasingη isdue tothejethavingtopassthrough moretrackermaterialbefore hittingtheECAL.

HighLevelTau TriggerusingthePixelDetector1 I

• DirectionoftheTau JetAxisisdeterminedbyjet

reconstructioninthecalorimeter.

• AMatchingConeofsolidangleRm aroundtheJetAxis

isdrawn.

• OnlytrackswithintheMatchingConewithtransverse

momentumgreaterthanpm

T areconsidered.

• Thetrackwiththehighesttransversemomentumamong

themiscalledtheLeadingTrack.

• ASignalConeofsolidangleRs aroundtheLeading

Trackisdrawn.

• AlltrackswithintheSignalConewhichhaveaz1

impactparameterlessthan<ztr awayfromthez1impact parameteroftheLeadingTrackareconsideredpartof theTau Jet.Theremustbe1or3suchtracks.

• AnIsolationConeofsolidangleRI isdrawnaroundthe

JetAxis.

• AlltrackswithintheIsolationConethathavez1impact

parameterlessthan<ztr awayfromthez1impact parameteroftheLeadingTrackandwithtransverse momentumgreaterthanpI

T areconsidered.Iftheyare

foundonlywithintheSignalCone,theisolationcriterion fortheTau Jetissatisfied.

HighLevelTau TriggerusingthePixelDetector1 II

PerformanceofHighLevelTau TriggerwithH/A/h→τ+(jet)τ1(jet)

TwoschemesforHLTonthetable,really.

• UsetheECAL+PixelDetector.Codename:.
• UsetheSiliconTracker.Codename: .

Calo+Pxl Algorithm

RateofeventscomingoutofLevel1issuppressedbyafactorof3usingECALIsolationwith Pisol

cut=5GeV/c ononeofthetwojets.

1. Trackerisolationisappliedtobothjets.Trackreconstructionrequiredforthisisdone entirelybythePixelSystem. 3pixelhitsissufficienttoreconstructtracks.Essentially,2 ofthe3hitsareusedfirst andmatchedinrΦ andzr toestablishtrackcandidates.Thematchingcutsare

• ptimised fortracksofpT~1GeV/c.Validpixelpairsarematchedwiththe3rd pixelhit

formingpixel1tracks.Themomentumofthesepixel1tracksisreconstructedwithouta primaryvertexconstraint.Alistofprimaryverticeswheretrackscrossthez1axisis drawnup.Primaryvertices(andtheirtracks)withatleast3tracksarekept,andthe restaredeleted. RM=0.1,pT

ltr=3GeV/c,RS=0.07,Ri=freeparameter.

2. Thefirstjetdefinesprimaryvertex.Alltracksofbothjetsmustoriginatethere,elsedeleted.

PerformanceofCalo+Pxl withH/A/h→τ+(jet)τ1(jet)

Efficiencyofacceptingsignaleventsversus efficiencyofacceptingQCDmulti1jetevents isplottedontheright.TheCalo1Pxl algorithmisappliedtothefirstjet.Two HiggsBosonmassesofMH=200and500 GeV/c2 areused.Ri isvariedfrom0.2to0.6 instepsof0.05 TheCalo1Pxl algorithmisnowappliedto bothjets.Thefirstjetisusedtodecidethe primaryvertexwherefromalltracksinboth jetsmustoriginate.QCDsuppressionis definitelyimproved.

H/A/h→τ+τ1→.,jetEventGenerationandPre1selection

• Signal
• H/A/h→τ+τ1 doneinPYTHIA
• τ+τ1→(.νν)(hadronic jet+ν)doneinTAUOLA

Background

• Drell Yan Z/γ*→τ+τ1 →(.νν)(hadronic jet+ν)producedintwobins:40<mττ <120

GeV/c2 andmττ >120GeV/c2

• ττbb wasproducedwith 60<mττ <100GeV/c2 andmττ >100GeV/c2
• W+jet wasproducedwithpT >20GeV/c
• tt
• Wt
• SUSYbackgroundnegligible
• MuonwasisolatedwithnochargedparticlewithpT >1GeV/c withinasolidangleof0.2
• Tau1Jetisolatedwithatmost1chargedtrackatsolidangle>0.1and<0.4.Leadingtrackinthis

jetrequiredtohavepT >3GeV/c.

H/A/h→τ+τ1→.,jetOfflineEventSelection

• Tau1Jetidentification.IsalmostidenticaltoCalo1Pxl identification.Except:
• insteadofhastilycreatedpixel1tracks,fullyreconstructedtracksareused.
• insteadofusingJetAxisdefinedbyECAL,sumofthemomentaofreconstructedtracksisused.
• theleadingtrackshouldhavepT >10GeV/c for1trackinsignalconeorpT >20GeV/c for3

tracksinsignalcone.

• btaggingofatleastoneb1JetisrequiredtomakesurewearedealingwithbbH(A).Oneb1jetwithpT

>20GeV/c wasrequired.

• Eventscontainingthedecayof(anearlyon1shell)Winto.+ν. arerejectedbyacutof60GeV/c2 on

theinvariantmassoftheW.

• CentralJetVeto:AlleventscontainingmorethanjusttheTau andbjetsin|η|<2.5withET>20GeV

arerejected.Reducestt background.

• EHCAL/pltr cut.
• Reconstructedmuonandτ1jetcannotbebacktoback.
• Eventswithnegativereconstructedneutrinoenergyrejected.

H/A/h→τ+τ1→.,jetSignalAnalysis

Theproductioncrosssectionsandindividualselection efficienciesforsignalsofMA=200and500GeV/c2

H/A/h→τ+τ1→.,jetBackgroundAnalysis

Theproductioncross sectionsandindividual selectionefficienciesfor backgroundprocesses.

H/A/h→τ+τ1→.,jetSystematicUncertainties

• NonZ/γ*backgroundeventshave12%systematicuncertaintyfromMonteCarlosimulation.
• Z/γ*
• Drell Yan processtoττ hassystematicerrorof8%tothenumberofevents.Contributionsfrom

b1tagginguncertaintyandjetscaleuncertainty.

• ττbbhas15%systematicuncertainty.

H/A/h→τ+τ1→.,jetHiggsReconstructionandDiscoveryReach

• 20fb11 worthofdatasimulated.LowmassHiggssignal

pronouncednear200GeV.Notsothe500GeV Higgs.

• 5σ discoveryareawithoutuncertaintyisborderedwiththe

solidcurve.Withsystematicuncertaintyborderedwith dashedcurve.

H/A/h→τ+τ1→e,jetEventGeneration

Signal

• H/A/h→τ+τ1 doneinPYTHIA
• τ+τ1→(.νν)(hadronic jet+ν)doneinTAUOLA

Background

• Z/γ*→τ+τ1 →e+hadronic jet+missingenergy
• bb Z/γ*,Z/γ*1>ττ 1>e+τ(jet)+missingenergy
• tt 1>bbW1W2,W1 1>τ(jet)+missingenergy,W2 1>e+missingenergyorW2 1>τ,e,

missingenergyandτ1>e+missingenergy

• W1t1>bW1W2,W1 1>τ+missingenergy,W2 1>e+missingenergyorW2 1>τ,e,missing

energyandτ1>e+missingenergy.(GeneratedwithTOPREX.) Thebackgroundsthatcanfakethesignalwhereahadronic jetoranelectroncanbeidentified asaτ1jetare:

• W+jet,W1>e+missingenergy
• Z/γ*1>e+e1
• bb1Z/γ*,Z/γ*1>e+e1
• tt 1>bbW1W2,W11>jet+jet,W21>e+missingenergy

H/A/h→τ+τ1→e,jetOfflineEventSelection

• Theisolatedelectronfromthedecayofoneofthetau leptonswasfirstsearchedfor.The

reconstructedelectronswerefirstrequiredtobeisolatedinthetrackerdemandingthatnotrack withpT >1GeV/c wasfoundwithinthesolid1angleR=0.4aroundtheelectrontrajectory.The ratioofhadronic clusterenergytoelectromagneticenergywasrequiredwascutwithTheratioof supercluster energytotrackmomentumwasgivenathreshold.Thepurityoftheselected electronswasfoundtobe97.5%.

• Thetau1jetidentificationwasidenticaltotheoneusedinττ 1>.+jetanalysis.
• Athresholdof0.35onwasusedtorejecthighlyenergeticelectronsthatmaybefaking

jets,similartothethresholdusedinττ 1>.+jetanalysis.

• Thechargesofthetau1jetandelectronwererequiredtobeopposite.
• Atleastonejetfromtheeventwasrequiredtobeab1jetwithET >20GeV and|η|<2.5.This

reducesthett background.Thisvetoefficiencywasfoundtobe60%forthesignaland5%inthe background.

• Eventswithelectronsandjetsbacktobackwererejectedwitha cutof

ltr HCAL

p E f =

( ) ( )

9962 . , cos − >

T T E

p

• ϕ

H/A/h→τ+τ1→e,jetSignalAnalysis

Theefficienciesofallthecutsonsignaleventswithtanβ=20andMA=130– 500GeV/c2 andthenumberofproducedeventsat30fb11 tricklingthroughthe selectioncriteriaas#ofevents(percentageefficiency).

H/A/h→τ+τ1→e,jetBackgroundAnalysis

At30fb11 ofevents,number

• fDrell Yan eventstrickling

throughselectioncriterion. At30fb11 ,otherbackground eventstricklingthroughthe selectioncriterion.

H/A/h→τ+τ1→e,jetSystematicUncertainties

Sourcesofuncertainty:

• Uncertaintyofelectronreconstructionandidentification~2%
• Uncertaintyincalorimeterenergyscalefortau reconstructiongaveuncertaintiesof:

Z/γ*1 5.1% bbZ/γ*1 3.8% tt – 7.3% Wt– 11.3% W+jet – 11.8%

• UncertaintyofcrosssectionsattheLHCare:

Z/γ*1 1% bbZ/γ*1 14.2% tt – 5.6% Wt– 5.6% W+jet – 5.6% Totaluncertaintiesinnumberofeventsforthebackgroundprocesses: Z/γ*1 8.1% bbZ/γ*1 15.9% tt – 11.1% Wt– 14.0% W+jet – 14.5%

H/A/h→τ+τ1→e,jetHiggsReconstruction

Thereconstructedinvariantττ massdistributionforMA=200GeV/c2,tanβ = 20andMA=300GeV/c2,tanβ =25for30fb11worthofdata.Thedashedline showsthesumoftheZ/γ*andbbZ/γ*backgrounds.

H/A/h→τ+τ1→e,jetDiscoveryReach

The 5σdiscoveryareasforthischannelofexplorationintheMA1tanβ planefor simulationswithoutandwithsystematicuncertaintieswith30fb11 ofdata.

H/A/h→τ+τ1→jet,jetEventGenerationandPre1selection

• Signal
• gg → H/A/h→τ+τ1 doneinPYTHIA
• τ+τ1→(.νν)(hadronic jet+ν)doneinTAUOLA

Background

• tt,Drell Yan productionofZ/γ*thatdecayintotau jets,W+jet,Wt– alldonein

PYTHIA.AllWandZwereforcedtodecayintotau1leptons.

• Hadronisation oftau doneusingTAUOLA.
• Theττbb backgroundwasgeneratedwithCOMPHEPandpropagatedtoPYTHIAfor

showeringandhadronisation anddecayoftau leptons.

• TheZ/γ*generationwassplitintothreebinsofinvariantmass 80<mττ <130GeV/c2,130

<mττ <300GeV/c2andmττ >300GeV/c2.

• TheW+jet backgroundwasgeneratedwithpT >65GeV/c.
• Alleventswerepre1selectedatgenerationleveltohavetwotau1likejets.Ajetisselectedas

tau1likeifithasET >50GeV,|η|<2.4andthetransversemomentumoftheleadingtrackpT >30GeV/c.

H/A/h→τ+τ1→jet,jetOfflineEventSelection

• ET >50GeV wasrequiredonbothtau1jetsforMA =200GeV/c2.ForhigherHiggsBoson

massesasymmetricalcutswereusedonthetwojets:100and50GeV forMA =500GeV/c2,150 and50GeV forMA=800GeV/c2.ThisisseentorejectQCDjetsmoreeffectively.

• Tau identificationwasdoneusingTrackerIsolationdescribedbefore.Theparametersusedwere

stricterthantheonesatHighLevelTriggerandareRm=0.1,RS=0.04,Ri=0.5,pT

ltr=35GeV/c and

pT

i=1GeV/c.OneorthreechargedtracksinthesignalconewererequiredforlowmassHiggs

whileeffectivebackgroundrejectionwaspossiblewiththestrictconditionofonlyonecharged trackinthesignalcone.Thetwotau1jetcandidateswererequiredtohaveoppositecharges.

• Atleastoneb1taggedjetwasrequiredintheevent.

H/A/h→τ+τ1→jet,jetSignalAnalysis

Theefficienciesatthetriggerandofflinelevelsofeventselection.Ntracks =1or3forMA =200GeV/c2 andNtracks =1forMA =500and800GeV/c2.

H/A/h→τ+τ1→jet,jetBackgroundAnalysis

TheQCDmulti1jet backgroundefficiencies. Group1:TheLevelOneand calorimetrictau1jet reconstructionatHighLevel andOffline. Group2:Tau1jetisolationat HLTandoffline. Group3:Extrab1taggedjet andHiggsmass reconstruction.

H/A/h→τ+τ1→jet,jetBackgroundAnalysis

Thenumberofexpectedeventswith60fb11andefficienciesoftheselectionsforthesomeof remainingreduciblebackgroundevents.

H/A/h→τ+τ1→jet,jetHiggsReconstruction

Thereconstructedinvariantmττ massesforthesignalsofMA =200GeV/c2 andtanβ =20whileFigure 18showsthesameforMA =500GeV/c2 andtanβ =30.Thethicksolidlineistheisolatedsignal.The normalsolidhistogramisthesignal+background.ThedashedhistogramistheQCD1jetbackground. Thesolid1dashedhistogramistheirreduciblebackground.

H/A/h→τ+τ1→jet,jetDiscoveryReach

The reachesof51σdiscoveryintheMA1tanβplanewithandwithoutsystematicuncertainties.