XXXI PHYSICS IN COLLISION, Vancouver, BC Canada, August 28 - September 1, 2011 ATLAS Jet Energy Scale D.Schouten2,A.Tanasijczuk1,M.Vetterli1,2,onbehalfoftheATLASCollaboration 1SimonFraserUniversityand2TRIUMF, Canada Jets originating from the fragmentation of quarks and gluons are the most common, and complicated, final stateobjectsproducedathadroncolliders. Apreciseknowledgeoftheirenergycalibrationisthereforeofgreat importance at experiments at the Large Hadron Collider at CERN, while is very difficult to ascertain. We presentin-situtechniquesandresultsforthejetenergyscaleatATLASusingrecentcollisiondata. ATLAShas demonstratedanunderstandingofthenecessaryjetenergycorrectionstowithin≈4%inthecentralregionof thecalorimeter. 2 1. Introduction > 0.8 P 1 E/ (0.0<|η|<0.6) < 0 0.7 Data 2010. L=866 µb 1 2 Pythia ATLAS MC10 The ATLAS experiment has been collecting colli- systematic uncertainty n sion data from the Large Hadron Coller (LHC) since 0.6 a early 2010. Currently, the ATLAS jet calibration is J 0.5 derived from Monte Carlo simulations, while its as- 1 sociated uncertainty is derived from a combination of 1 0.4 single hadron and dijet response measurements, and x] systematic variations in Monte Carlo simulations. In 0.3 ATLASPreliminary ordertovalidatethisapproach,ATLAShasemployed e A - a number of approaches to demonstrate an under- DAT 1.1 p standing of the jet energy scale. C/ M e In Section 1 the single particle response measure- 1 h [ ment in the central barrel region is presented. This 0.9 is extrapolated to the endcap regions of the calorime- 1 terusingthedijectrelativeresponsemeasurement, as 1 10 p[GeV] v described in Section 2. Section 3 details the photon 9 + jet measurements. Finally, in Section 4 a summary 2 Figure 1: Comparison of E/p measured in data (black 4 of all of the validation methods for the ATLAS jet points) and Monte Carlo simulations (green blocks) [1]. 2 energy scale (JES) is presented. The ratio of data to MC is also shown, in which the grey . 1 blocks represent the systematic and the bars represent 0 the statistical uncertainty. 2 1 2. Single Particle Response : v i 3. Relative Response X r The basic idea underyling the single particle re- a sponse is to measure the calorimeter response for iso- latedsingleparticlesbycomparingtheenergyandmo- mentum(tracking)measurements,namelyE/p,under For jets outside of the central barrel, the response theassumptionthatthetrackingmeasurementisvery for the central region is extrapolated using a dijet precise. Uncertainties for single particles are derived balance technique [2]. This procedure measures a re- from deviations of this measurement in simulations sponseforajetrelativetothecentralregionunderthe comparedtodata. Then,theseareextrapolatedtojet assumptionofmomentumbalanceofthedijetsystem, uncertainties using simulations. Although the trans- andcomparesittotheresultinsimulations. TheJES lation from single particles to the jet context is non- uncertainty in the endcap region is then the sum in trivial, it has been exhaustively cross-checked and is quadrature of the uncertainties in the central region foundtohavesmalluncertainty. Forchargedparticles and the dijet relative response measurements. Cur- in the momentum range 0.5 < p < 20 GeV the E/p rently, the latter component (shown in Figure 2) is measuredinsituisusedtodeterminetheresponse[1]. the dominant one in the forward region, due to a dis- A comparison of E/p in data to that in Monte Carlo agreement between different Monte Carlo generators is shown in Figure 1. in the modelling of the reference balance assumption. 47 XXXI PHYSICS IN COLLISION, Vancouver, BC Canada, August 28 - September 1, 2011 1.15 C RM MPF EM, EMJES all jet algorithms /a Uncertainty Dat 1.1 Statistical R Systematic Total 1.05 1 ∫L dt = 38 pb 1 0.95 s = 7 TeV ATLAS Preliminary Figure 2: Systematic uncertainty derived for the relative correction using dijet balance as a function of jet p for 0.9 T 0 100 200 various regions of pseudorapidity [2]. pγ [GeV] T 4. Response from Photon + Jet Events Figure 3: Ratio of MPFDATA/MPFMC versus photon p [3]. The total uncertainty is indicated by the magenta T band, while the systematic uncertainty only is shown by Becausephotonsarewell-measuredobjects,onecan the yellow band. directly measure the jet response by using the princi- ple of momentum balance between a photon and re- coiljetinphoton+jetevents. Onetechnique, known for the recoil jets using the standard approaches de- as the missing ET projection fraction (MPF) directly scribed above, since they are in a reachable pT range measuresthetotalcalorimeterresponsetojetsbybal- for these methods. Then the momentum balance of ancing the hadronic recoil against the photon. The the recoil system and lead jet can be compared in MPF equation is: data and simulations. Bycombiningthesetechniqueswiththosedescribed Emiss·nˆ above, a robust validation of the JES and its uncer- R =1+ T γ. (1) MPF pγ tainty can be shown [6]. This is summarized in the T central region in Figure 4. Directly balancing the photon and jet in these events Insummary,thejetenergyscaleanditsuncertainty is a complementary technique, and is differently sen- inATLAS,derivedfromMonteCarlosimulations,has sitive to radiative effects [3]. A comparison of the been extensively validated to within ±4% in the cen- MPF in data and simulation is shown in Figure 3. tral region of the calorimeter for pT > 20 GeV (or The Monte Carlo simulation agrees with the data to ±2.5% for pT > 40 GeV) using a variety of comple- within a few percent over the entire range of photon mentary approaches. p . T C1.14 M1.12 5. Summary of Jet Energy Scale Data / 1.10.81 MTγ r uajelctitk jd ejietretct balance ATLASPreliminary 1.06 γ jet MPF 1.04 Besides the techniques summarized above, compar- 1.02 ing track and calorimeter jets and also measuring 1 transversemomentumbalanceinmulti-jetfinalstates 0.98 are useful probes of the jet energy scale. 0.96 0.94 The track jet comparison test works off of the as- 0.92 JES uncertainty anti k R=0.6, EM+JES sumptionthattheratioofthechargeparticlemomen- t 0.9 tumtothetotaljetmomentumistightlyconstrained. 102 103 pjet [GeV] Thus, by directly measuring the ratio of the track jet T tothematchedcalorimeterjetmomentumindataand Figure 4: Summary of all the jet energy scale in simulations, the validity of the simulation can be measurements performed by ATLAS, for the central determined [4]. calorimeter region [6]. The shaded graph is the Employing momentum balance in multi-jet final uncertainty for the Monte Carlo simulation-based jet states in which a high pT jet recoils against many calibration. lower p jets allows for validation at very high p [5]. T T This is because the uncertainties can be ascertained 47 XXXI PHYSICS IN COLLISION, Vancouver, BC Canada, August 28 - September 1, 2011 References [4] The ATLAS Collaboration, ATLAS-CONF- 2011-067, cdsweb.cern.ch/record/1349308 (2011) [1] The ATLAS Collaboration, ATLAS-CONF- [5] The ATLAS Collaboration, ATLAS-CONF- 2010-052, cdsweb.cern.ch/record/1281309 (2010) 2011-029, cdsweb.cern.ch/record/1337076 (2011) [2] The ATLAS Collaboration, ATLAS-CONF- [6] The ATLAS Collaboration, ATLAS-CONF- 2011-014, cdsweb.cern.ch/record/1334876 (2011) 2011-032, cdsweb.cern.ch/record/1337781 (2011) [3] The ATLAS Collaboration, ATLAS-CONF- 2011-031, cdsweb.cern.ch/record/1337781 (2011) 47