subtracting the underlying-event energy in reconstructed

1
Figure 2 (Above): The jet-energy spectrum with UE energy subtracted via methods 2 and 3. Also shown are the Gaussian fits to the negative jet energies to estimate the combinatorial jet contributions. [1] EPJC (2012) 72:1896 [2] Z. Chang’s thesis, “Inclusive Jet Longitudinal Double-Spin Asymmetry All Measurements in 510 GeV Polarized pp Collisions at STAR” (2016) and ALICE (perpendicular cones method) PRD 91, 112012(2015) All 3 methods of subtracting background in the reconstructed charged jet spectrum agree within a few percent for energies greater than 1.5 GeV and within ~10% for energies greater than 1 GeV. Methods 2 and 3, which first subtract UE energy, and subsequently the combinatorial jet component, agree to within ~25% to the lowest jet energies measured. Since Method 1 assumes no UE energy, subtracting only combinatorial jets as a background, we can conclude that this assumption is good for R=0.3 and jet energy greater than 1 GeV. . Subtracting the Underlying-Event Energy in Reconstructed Jets in = 200 GeV Proton+Proton Collisions at STAR C. Ostberg Department of Physics, San Francisco State University REU Student at the Cyclotron Institute, Texas A&M Mentored by Professor Saskia Mioduszewski The Anti-kt algorithm is used to cluster the jets. This algorithm picks a particle in a region, and clusters other particles in near vicinity, in terms of azimuth (Δφ) and pseudorapidity (Δη), to form a roughly conical jet. %& = 1 max -% . , -& . %& . . %3 = 1 -% . π-3π/4 π+3π/4 π-π/2 π+π/2 Recoil Side Uncorrelated jet candidates Uncorrelated jet candidates [UE] [UE] Figure 3 (Above): The background-subtracted jet-energy spectra for all 3 methods. The insert shows a zoomed-in view of the low jet-energy q Collimated spray of Hadrons q Proton Proton two back to back jets Collimated spray of Hadrons In ultra-relativistic collisions, the energy is high enough for “hard” scatterings to occur. These are characterized by a large momentum transfer between partons of the colliding baryons/nuclei. The hard-scattered partons then fragment into collimated sprays of hadrons, called “jets”. These jets are used to probe the QGP in heavy ion collisions. Not all of the energy from the collision goes towards the hard scattered jets. Uncorrelated soft particles are also produced, and contribute to the observed hard jet energy. These particles are what the background is composed of, and what must be subtracted. The figure below defines the “recoil” region (Δφ = π ± π/4 with respect to the trigger particle) and the “UE” region (underlying-event region where Δφ = ±π/2 − π/4 with respect to the trigger particle). There are two types of background: 1.) “combinatorial” jets – those clustered by the jet-finder, but not originating from hard scatterings, and 2.) underlying-event energy included in the clustering of the true “hard” jets. For R=0.3 (relatively small jets), the latter is expected to be small, but how small was studied here. Method 3 is also a jet by jet subtraction, but is more specific with how it calculates the average 5 . Circular areas with jet parameter R are created at +/- /2, with respect to the recoil jet’s , and at the same η as the recoil jet. There, the average 5 per area is found for each region ( ; and < ). These quantities are then averaged ( =>? ) and multiplied by the area of the recoil jet in order to scale it to the individual jets size ( 5 @A ). Combinatorial jets are then subtracted as follows. The background-jet spectra in Fig. 1 fall much faster than the recoil-region jet spectrum, showing that the overall background energy is small, and the combinatorial jets are typically made up of a small number of soft (low energy) particles. Methods 2 and 3 result in similar UE-energy subtracted spectra (Fig. 2). The combinatorial jets are subtracted assuming the Gaussian distribution (fit to the negative jet energies). For jet energies greater than 1.5 GeV, all three methods agree in the background-subtracted results shown in Fig. 3. Introduction We control the area of the jet by changing the jet parameter R (or jet radius) in the equation above. . + ∆ . (η = ln( <L M . ) is the angle between the jet and beam axis, = is the azimuthal angle in phase space). In this study R = 0.3 radians, and only charged particles are included in jet reconstruction. Data Used to Reconstruct Jets: STAR Experiment data from 2009 RHIC Run. = 200 GeV proton-proton collisions Events with trigger particle of 9 < 5 < 30 GeV Charged tracks with 0.2 < 5 < 20 GeV Recoil Jet 5 Average UE Jet 5 Off Cone 5 (Method 3) UE Jet 5 (Method 1) 5,&?- Q?RS [GeV] Jet-Reconstruction Jets resulting from hard scatterings (i.e. scatterings with large momentum transfer) provide insight into parton energy loss in the hot, dense medium produced by ultra-relativistic heavy-ion collisions. Complementary measurements in the medium-free proton-proton environment establish a vacuum fragmentation reference. In a collision, energy that goes into the production of particles not originating from the hard scattering is background (referred to as the underlying event) that must be subtracted from the measured jet-energy. We present a study of different methods to subtract the underlying- event energy in 200 GeV proton-proton collisions recorded at the STAR experiment at RHIC. Abstract Results Acknowledgements References Summary Methods < 5,&?- UV >= 5,% UV Y %Z[ &?-] [1] [2] 5,&?- ^_@ [GeV] Method 3 (After Gaussian Method 1 Method 2 (After Gaussian Sub) Figure 1 (Above): The raw recoil jet energy distribution (blue) before subtracting the backgrounds, as well as the energy distributions of the three background estimates Method 1 uses reconstructed jets in the UE region (the two Uncorrelated jet regions shown above) and subtracts their 5 (transverse momentum) statistically from the recoil jets’ 5 . This procedure subtracts an estimate of combinatorial jets, assuming the UE background is small enough to neglect. 5,&?- ^_@ [GeV] Method 3 (Off-Cone Sub) Method 3 Gaussian Fit Method 2 (Avg UE Subtraction) Method 2 Gaussian Fit Method 2 uses the average 5 (< 5,&?- UV >) of the same UE jets, but instead subtracts it on a jet-by-jet basis from the recoil jets. This procedure subtracts an estimate of the UE energy, leaving combinatorial jets as a remaining background. These will be subtracted using a separate procedure described below. NSF Grant (PHY – 1659847) DOE Grant (DE-FG02-07ER41485) Thank you to Dr. Nihar Sahoo and Professor Saskia Mioduszewski for mentoring me and allowing me to experience what its like to be a nuclear physicist for a summer. P p trk T A con = σ - P p trk T A con = σ + σ ave = 1 2 (σ - + σ + ) p bg T = σ ave A jet Combinatorial Jet Subtraction for Methods 2 and 3 Since negative jet energies are unphysical, we take entries in these bins to be an estimate of the contribution from combinatorial jets. Assuming the combinatorial jets show up as a symmetric distribution around zero energy (after UE energy subtraction), we fit a Gaussian function to the negative side of the distribution, reflect it to the positive side, and subtract the full Gaussian from the jet energy distribution. p+p 200 GeV (Run 9) p+p 200 GeV (Run 9) p+p 200 GeV (Run 9)

Upload: others

Post on 28-Feb-2022

1 views

Category:

Documents


0 download

TRANSCRIPT

Figure2(Above):Thejet-energyspectrumwithUEenergysubtractedviamethods2and3. AlsoshownaretheGaussianfitstothenegativejetenergiestoestimatethecombinatorialjetcontributions.

• [1]EPJC(2012)72:1896• [2]Z.Chang’sthesis,“InclusiveJetLongitudinalDouble-SpinAsymmetryAllMeasurementsin

510GeVPolarizedppCollisionsatSTAR”(2016)andALICE(perpendicularconesmethod)PRD91,112012(2015)

• All3methodsofsubtractingbackgroundinthereconstructedchargedjetspectrumagreewithinafewpercentforenergiesgreaterthan1.5GeVandwithin~10%forenergiesgreaterthan1GeV.• Methods2and3,whichfirstsubtractUEenergy,andsubsequentlythecombinatorialjetcomponent,agreetowithin~25%tothelowestjetenergiesmeasured.• SinceMethod1assumesnoUEenergy,subtractingonlycombinatorialjetsasabackground,wecanconcludethatthisassumptionisgoodforR=0.3andjetenergygreaterthan1GeV.

• .

SubtractingtheUnderlying-EventEnergyinReconstructedJetsin√𝐬𝐍𝐍 =200GeVProton+ProtonCollisionsatSTAR

C.OstbergDepartmentofPhysics,SanFranciscoStateUniversityREUStudentattheCyclotronInstitute,TexasA&M

MentoredbyProfessorSaskiaMioduszewski

TheAnti-kt algorithmisusedtoclusterthejets.Thisalgorithmpicksaparticleinaregion,andclustersotherparticlesinnearvicinity,intermsofazimuth(Δφ)andpseudorapidity (Δη),to

formaroughlyconicaljet.

𝑑%& =1

max 𝑝-%. , 𝑝-&.∆𝑅%&.

𝑅. 𝑑%3 =1𝑝-%.

π-3π/4

π+3π/4

π-π/2

π+π/2

Recoil Side

Uncorrelated jet candidates

Uncorrelated jetcandidates

[UE]

[UE]

Figure3(Above):Thebackground-subtractedjet-energyspectraforall3methods.Theinsertshowsazoomed-inviewofthelowjet-energy

q

Collimated spray of Hadrons

q

Proton

Proton

two back to back jets

Collimated spray of Hadrons

Inultra-relativisticcollisions,theenergyishighenoughfor“hard”scatteringstooccur. Thesearecharacterizedbyalargemomentumtransferbetweenpartons ofthecollidingbaryons/nuclei.Thehard-scatteredpartons thenfragmentintocollimatedspraysofhadrons,called“jets”.These

jetsareusedtoprobetheQGPinheavyioncollisions.

Notalloftheenergyfromthecollisiongoestowardsthehardscatteredjets.Uncorrelatedsoftparticlesarealsoproduced,andcontributetotheobservedhardjetenergy.Theseparticlesarewhatthebackgroundiscomposedof,and

whatmustbesubtracted.

Thefigurebelowdefinesthe“recoil”region(Δφ =π± π/4withrespecttothetriggerparticle)andthe“UE”region(underlying-eventregionwhereΔφ =±π/2−π/4withrespecttothetriggerparticle).Therearetwotypesofbackground:1.)“combinatorial”jets– thoseclusteredbythejet-finder,butnotoriginatingfromhardscatterings,and2.)underlying-eventenergyincludedintheclusteringofthetrue“hard”jets. ForR=0.3(relativelysmalljets),thelatterisexpectedtobesmall,buthowsmallwasstudied

here.

Method3isalsoajetbyjetsubtraction,butismorespecificwithhowitcalculatestheaverage𝑃5.CircularareaswithjetparameterRarecreatedat𝜑 +/- 𝜋/2,withrespecttotherecoiljet’s𝜑,andatthesameη astherecoiljet.There,theaverage𝑃5 perareaisfoundforeachregion(𝜎; and𝜎<).Thesequantitiesarethenaveraged(𝜎=>?)andmultipliedbythearea

oftherecoiljetinordertoscaleittotheindividualjetssize(𝑃5@A).

Combinatorialjetsarethensubtractedasfollows.

• Thebackground-jetspectrainFig.1fallmuchfasterthantherecoil-regionjetspectrum,showingthattheoverallbackgroundenergyissmall,andthecombinatorialjetsaretypicallymadeupofasmallnumberofsoft(lowenergy)particles.

• Methods2and3resultinsimilarUE-energysubtractedspectra(Fig.2).ThecombinatorialjetsaresubtractedassumingtheGaussiandistribution(fittothenegativejetenergies).• Forjetenergiesgreaterthan1.5GeV,allthreemethodsagreeinthebackground-subtractedresultsshowninFig.3.

Introduction

WecontroltheareaofthejetbychangingthejetparameterR(orjetradius)intheequationabove.𝑅 ≈ ∆𝜑. + ∆𝜂.� (η =ln(𝑡𝑎𝑛<L M

.) ,θ

istheanglebetweenthejetandbeamaxis,𝜑 =istheazimuthalangleinphasespace).Inthis

studyR=0.3radians,andonlychargedparticlesareincludedinjetreconstruction.

DataUsedtoReconstructJets:• STARExperimentdatafrom2009RHICRun.• √𝒔𝑵𝑵 =200GeVproton-protoncollisions• Eventswithtriggerparticleof9<𝑃5 <30GeV• Chargedtrackswith0.2<𝑃5 <20GeV

RecoilJet𝑃5AverageUEJet𝑃5(Method2)OffCone𝑃5 (Method3)UEJet𝑃5 (Method1)

𝑃5,&?-Q?RS [GeV]

Jet-Reconstruction

Jetsresultingfromhardscatterings(i.e.scatteringswithlargemomentumtransfer)provideinsightintoparton energylossinthehot,densemediumproducedbyultra-relativisticheavy-ioncollisions.Complementarymeasurementsinthemedium-freeproton-protonenvironmentestablishavacuumfragmentationreference.Inacollision,energythatgoesintotheproductionofparticlesnotoriginatingfromthehardscatteringisbackground(referredtoastheunderlyingevent)thatmustbesubtractedfromthemeasuredjet-energy.Wepresentastudyofdifferentmethodstosubtracttheunderlying-eventenergyin200GeVproton-protoncollisionsrecordedattheSTARexperimentatRHIC.

Abstract

Results

Acknowledgements References

Summary

Methods

< 𝑃5,&?-UV >=∑ 𝑃5,%UVY%Z[

𝑁&?-]

[1]

[2]

𝑃5,&?-^_@ [GeV]

Method3(AfterGaussianSub)

Method1

Method2(AfterGaussianSub)

Figure1(Above):Therawrecoiljetenergydistribution(blue)beforesubtractingthebackgrounds,aswellastheenergydistributionsofthethreebackgroundestimates

Method1usesreconstructedjetsintheUEregion(thetwoUncorrelatedjetregionsshownabove)andsubtractstheir𝑃5 (transversemomentum)statisticallyfromtherecoiljets’𝑃5 .Thisproceduresubtractsanestimateofcombinatorialjets,assumingtheUEbackgroundissmallenoughto

neglect.

𝑃5,&?-^_@ [GeV]

Method3(Off-Cone Sub)Method3GaussianFitMethod2(AvgUESubtraction)Method2GaussianFit

Method2usestheaverage𝑃5 (< 𝑃5,&?-UV >)ofthesameUEjets,butinsteadsubtractsitonajet-by-jetbasisfromtherecoiljets.ThisproceduresubtractsanestimateoftheUEenergy,leavingcombinatorialjetsasaremainingbackground. Thesewillbesubtractedusingaseparate

proceduredescribedbelow.

• NSFGrant(PHY– 1659847)• DOEGrant(DE-FG02-07ER41485)• ThankyoutoDr.Nihar Sahoo andProfessorSaskiaMioduszewskiformentoringme

andallowingmetoexperiencewhatitsliketobeanuclearphysicistforasummer.

PptrkT

Acon

= ��

PptrkT

Acon

= �+

�ave =1

2(�� + �+)

pbgT = �ave ⇥Ajet

CombinatorialJetSubtractionforMethods2and3Sincenegativejetenergiesareunphysical,wetakeentriesinthesebinstobeanestimateofthecontributionfromcombinatorialjets. Assumingthe combinatorialjetsshowupasasymmetricdistributionaroundzeroenergy(afterUEenergysubtraction),wefitaGaussianfunctiontothe

negativesideofthedistribution,reflectittothepositiveside,andsubtractthefullGaussianfromthejetenergydistribution.

p+p 200GeV(Run9)

p+p 200GeV(Run9)

p+p 200GeV(Run9)