Jet energy measurement and its systematic uncertainty in proton proton collisions at s = 7 TeV with the ATLAS detector

Transcription

1 Eur. Phys. J. C (5) 75:7 DOI.4/ejc/s5-4-9-y Regular Article - Exerimental Physics Jet energy measurement and its systematic uncertainty in roton roton collisions at s = 7 ev with the ALAS detector ALAS Collaboration CERN, Geneva, Switzerland Received: June 4 / Acceted: 4 November 4 CERN for the benefit of the ALAS collaboration 4. his article is ublished with oen access at Sringerlink.com Abstract he energy scale (JES) and its systematic uncertainty are determined for s measured with the ALAS detector using roton roton collision data with a centre-ofmass energy of s = 7 ev corresonding to an integrated luminosity of 4.7 fb. Jets are reconstructed from energy deosits forming toological clusters of calorimeter cells using the anti-k t algorithm with distance arameters R =.4 or R =.6, and are calibrated using MC simulations. A residual JES correction is alied to account for differences between data and MC simulations. his correction and its systematic uncertainty are estimated using a combination of in situ techniques exloiting the transverse momentum balance between a and a erence object such as a hoton or a Z boson, for < GeV and seudoraidities η < 4.5. he effect of multile roton roton interactions is corrected for, and an uncertainty is evaluated using in situ techniques. he smallest JES uncertainty of less than % is found in the central calorimeter region ( η <.) for s with 55 < 5 GeV. For central s at lower, the uncertainty is about %. A consistent JES estimate is found using measurements of the calorimeter resonse of single hadrons in roton roton collisions and test-beam data, which also rovide the estimate for > ev. he calibration of forward s is derived from di balance measurements. he resulting uncertainty reaches its largest value of 6 % for low- s at η =4.5. Additional JES uncertainties due to secific event toologies, such as close-by s or selections of event samles with an enhanced content of s originating from light quarks or gluons, are also discussed. he magnitude of these uncertainties deends on the event samle used in a given hysics analysis, but tyically amounts to.5 %. atlas.ublications@cern.ch Contents Introduction... he ALAS detector.... Detector descrition.... Calorimeter ile-u sensitivity... Monte Carlo simulation of s in the ALAS detector.... Inclusive Monte Carlo simulation samles.. Z- and γ - Monte Carlo simulation samles.... o-quark air Monte Carlo simulation samles....4 Minimum bias samles....5 Detector simulation... 4 Dataset... 5 Jet reconstruction and calibration with the ALAS detector oological clusters in the calorimeter Jet reconstruction and calibration Jet quality selection rack s ruth s Jet kinematics and directions... 6 Jet energy correction for ile-u interactions Pile-u correction method Princial ile-u correction strategy Derivation of ile-u correction arameters Pile-u validation with in situ techniques and effect of out-of-time ile-u in different calorimeter regions... 7 In situ transverse momentum balance techniques.. 7. Relative in situ calibration between the central and forward raidity regions In situ calibration methods for the central raidity region... 8 Relative forward- calibration using di events.. 8. Intercalibration using events with di toologies...

2 7 Page of Eur. Phys. J. C (5) 75:7 8.. Intercalibration using a central erence region Intercalibration using the matrix method Event selection for di analysis rigger selection Dataset and quality selection Di toology selection Di balance results Binning of the balance measurements Comarison of intercalibration methods 8.. Comarison of data with Monte Carlo simulation Derivation of the residual correction Systematic uncertainty Modelling uncertainty Sub-leading radiation suression φ(, ) event selection rigger efficiencies Imact of ile-u interactions Jet resolution uncertainty Summary of the η-intercalibration and its uncertainties... 9 Jet energy calibration using Z- events Descrition of the balance method Selection of Z- events Measurement of the balance Measuring out-of-cone radiation and underlying event contributions Systematic uncertainties Fitting rocedure Extraolation rocedure Additional radiation suression Out-of-cone radiation and underlying event Imact of additional ile-u interactions Electron energy scale Imact of the Monte Carlo generator Summary of systematic uncertainties. 9.6 Summary of the Z- analysis... Jet energy calibration using γ - events.... In situ calibration techniques.... Event selection of γ - events.... Jet resonse measurement....4 Systematic uncertainties of hoton balance Influence of ile-u interactions Soft-radiation suression Background from events Photon energy scale Jet energy resolution Monte Carlo generator Out-of-cone radiation and underlying event Summary of systematic uncertainties..5 Summary of the γ - analysis... High- energy calibration using multi events.... Multi balance technique and uncertainty roagation.... Selection of multi events.... Multi balance measurement....4 Systematic uncertainties on the multi balance....5 Summary of multi analysis... Forward- energy measurement validation using Z- and γ - data... Jet energy calibration and uncertainty combination.... Overview of the combined JES calibration rocedure.... Combination technique.... Uncertainty sources of the in situ calibration techniques....4 Combination results....5 Comarison of the γ - calibration methods..6 Simlified descrition of the correlations....7 Jet energy scale correlation scenarios....8 Alternative reduced configurations Comarison to energy scale uncertainty from single-hadron resonse measurements Jet energy scale uncertainty from the W boson mass constraint Event samles Reconstruction of the W boson Extraction of the relative light scale Systematic uncertainties Results... 6 Systematic uncertainties on corrections for ile-u interactions Event and object selection Derivation of the systematic uncertainty Summary on ile-u interaction corrections.. 7 Close-by effects on energy scale Samles and event selection Non-isolated energy scale uncertainty... 8 Jet resonse difference for quark and gluon induced s and associated uncertainty Event selection Jet and track selection Jet flavour definition Dataset for flavour studies Calorimeter resonse to quark and gluon induced s...

3 Eur. Phys. J. C (5) 75:7 Page of 7 8. Discrimination of light-quark and gluon induced s Summary of the flavour deendence analysis... 9 Jets with heavy-flavour content Jet selection and resonse definition rack selection Event selection Jet samle selection o-quark air samle selection MC-based systematic uncertainties on the calorimeter b- energy scale Calorimeter energy measurement validation using tracks Systematic uncertainties Results Semiletonic correction and associated uncertainties Semiletonic neutrino energy validation using di balance Conclusions on heavy-flavour s... Jet resonse in roblematic calorimeter regions.... Correction algorithms for non-oerating calorimeter modules..... Correction based on calorimeter cell energies..... Corrections based on shaes.... Performance of the bad calorimeter region corrections..... Conclusion on bad calorimeter regions... Summary of the total energy scale systematic uncertainty... Conclusions... Acknowledgments... Aendix A: Comarison of the ALAS JES uncertainty with revious calibrations... References... Introduction Jets are the dominant feature of high-energy, hard roton roton interactions at the Large Hadron Collider (LHC) at CERN. hey are key ingredients of many hysics measurements and for searches for new henomena. In this aer, s are observed as grous of toologically related energy deosits in the ALAS calorimeters, associated with tracks of charged articles as measured in the inner tracking detector. hey are reconstructed with the anti-k t algorithm [] and are calibrated using Monte Carlo (MC) simulation. A first estimate of the energy scale (JES) uncertainty of about 5 9 % deending on the transverse momentum ( ), described in Ref. [], is based on information available before the first roton roton collisions at the LHC, and initial roton roton collision data taken in. A reduced uncertainty of about.5 % in the central calorimeter region over a wide range of 6 < 8 GeV was achieved after alying the increased knowledge of the detector erformance obtained during the analysis of this first year of ALAS data taking []. his estimation used single-hadron calorimeter resonse measurements, systematic variations of MC simulation configurations, and in situ techniques, where the transverse momentum is comared to the of a erence object. hese measurements were erformed using the dataset, corresonding to an integrated luminosity of 8 b [4]. During the year the ALAS detector [5] collected roton roton collision data at a centre-of-mass energy of s = 7 ev, corresonding to an integrated luminosity of about 4.7 fb. he larger dataset makes it ossible to further imrove the recision of the energy measurement, and also to aly a correction derived from detailed comarisons of data and MC simulation using in situ techniques. his document resents the results of such an imroved calibration of the energy measurement and the determination of the uncertainties using the dataset. he energy measurement of s roduced in rotonroton and electron-roton collisions is also discussed by other exeriments [6 7]. he outline of the aer is as follows. Section describes the ALAS detector. he Monte Carlo simulation framework is resented in Sect., and the used dataset is described in Sect. 4. Section 5 summarises the reconstruction and calibration strategy. he correction method for the effect of additional roton roton interactions is discussed in Sect. 6. Section 7 rovides an overview of the techniques based on balance that are described in detail in Sects. 8 to. First the intercalibration between the central and the forward detector using events with two high- s is resented in Sect. 8. hen, in situ techniques to assess differences of the energy measurement between data and Monte Carlo simulation exloiting the balance between a and a wellmeasured erence object are detailed. he erence objects are Z bosons in Sect. 9, hotons in Sect., and a system of low- s in Sect.. he validation of the forward- energy measurements with balance methods using Z- and γ - events follows in Sect.. he strategy on how to extract a final calibration out of the combination of in situ techniques, and the evaluation strategies for determining the corresonding systematic uncertainties, are discussed in Sect.. he same section also shows the final result of the calibration, including its systematic uncertainty, from the combination of the in situ techniques. Section 4 comares the JES uncertainty as derived from the single-hadron calorimeter resonse measurements to that

4 7 Page 4 of Eur. Phys. J. C (5) 75:7 obtained from the in situ method based on balance discussed in the receding sections. Comarisons to JES uncertainties using the W boson mass constraint in final states with hadronically decaying W bosons are resented in Sect. 5. Additional contributions to the systematic uncertainties of the measurement in ALAS are resented in Sects. 6 8, where the correction for the effect of additional roton roton interactions in the event, the resence of other closeby s, and the resonse deendence on the fragmentation ( flavour) are discussed. he uncertainties for exlicitly tagged s with heavy-flavour content are outlined in Sect. 9. A brief discussion of the correction of the calorimeter energy in regions with hardware failures and the associated uncertainty on the energy measurement is resented in Sect.. A summary of the total energy scale uncertainty is given in Sect.. Conclusions follow in Sect.. A comarison of the systematic uncertainties of the JES in ALAS with revious calibrations is resented in Aendix A. he ALAS detector. Detector descrition he ALAS detector consists of a tracking system (Inner Detector, or ID in the following), samling electromagnetic and hadronic calorimeters and muon chambers. A detailed descrition of the ALAS exeriment can be found in Ref. [5]. he Inner Detector has comlete azimuthal coverage and sans the seudoraidity region η <.5. It consists of layers of silicon ixel detectors, silicon microstri detectors and transition radiation tracking detectors, all of which are immersed in a solenoid magnet that rovides a uniform magnetic field of. Jets are reconstructed using the ALAS calorimeters, whose granularity and material varies as a function of η. he electromagnetic calorimetry (EM) is rovided by highgranularity liquid-argon samling calorimeters (LAr), using lead as an absorber. It is divided into one barrel ( η <.475) and two end-ca (.75 < η <.) regions. he hadronic calorimetry is divided into three distinct sections. he most central contains the central barrel region ( η <.8) and two extended barrel regions (.8 < η <.7). hese regions are instrumented with scintillator-tile/steel hadronic ALAS uses a right-handed coordinate system with its origin at the nominal interaction oint (IP) in the centre of the detector and the z-axis along the beam ie. he x-axis oints from the IP to the centre of the LHC ring, and the y-axis oints uward. Cylindrical coordinates (r,φ) are used in the transverse lane, φ being the azimuthal angle around the beam ie. he seudoraidity is defined in terms of the olar angle θ as η = ln tan(θ/). calorimeters (ile). Each barrel region consists of 64 modules with individual φ coverages of. rad. he two hadronic end-ca calorimeters (HEC;.5 < η <.) feature liquid-argon/coer calorimeter modules. he two forward calorimeters (FCal;. < η < 4.9) are instrumented with liquid-argon/coer and liquid-argon/tungsten modules to rovide electromagnetic and hadronic energy measurements, resectively. he muon sectrometer surrounds the ALAS calorimeter. A system of three large air-core toroids, a barrel and two endcas, generates a magnetic field in the seudoraidity range of η <.7. he muon sectrometer measures muon tracks with three layers of recision tracking chambers and is instrumented with searate trigger chambers. he trigger system for the ALAS detector consists of a hardware-based Level (L) and a software-based High Level rigger (HL) [8]. At L, s are first built from coarse-granularity calorimeter towers using a sliding window algorithm, and then subjected to early trigger decisions. his is ined using s reconstructed from calorimeter cells in the HL, with algorithms similar to the ones alied offline.. Calorimeter ile-u sensitivity One imortant feature for the understanding of the contribution from additional roton roton interactions (ile-u) to the signal in the dataset is the sensitivity of the ALAS liquid argon calorimeters to the bunch crossing history. In any LAr calorimeter cell, the reconstructed energy is sensitive to the roton roton interactions occurring in aroximately ( data, 4 at LHC design conditions) receding and one immediately following bunch crossings (out-oftime ile-u), in addition to ile-u interactions in the current bunch crossing (in-time ile-u). his is due to the relatively long charge collection time in these calorimeters (tyically 4 6 ns), as comared to the bunch crossing intervals at the LHC (design 5 ns and actually 5 ns in data). o reduce this sensitivity, a fast, biolar shaed signal is used with net zero integral over time. he signal shaes in the liquid argon calorimeters are otimised for this urose, leading to cancellation on average of in-time and out-of-time ile-u in any given calorimeter cell. By design of the shaing amlifier, the most efficient suression is achieved for 5 ns bunch sacing in the LHC beams. It is fully effective in the limit where, for each bunch crossing, about the same amount of energy is deosited in each calorimeter cell. he beam conditions, with 5 ns bunch sacing and a relatively low cell occuancy from the achieved instantaneous luminosities, do not allow for full ile-u suression he shaed ulse has a duration exceeding the charge collection time.

5 Eur. Phys. J. C (5) 75:7 Page 5 of 7 by signal shaing, in articular in the central calorimeter region. Pile-u suression is further limited by large fluctuations in the number of additional interactions from bunch crossing to bunch crossing, and in the energy flow atterns of the individual collisions in the time window of sensitivity of aroximately 6 ns. Consequently, the shaed signal extracted by digital filtering shows a rincial sensitivity to in-time and out-of-time ile-u, in articular in terms of a residual non-zero cell-signal baseline. his baseline can lead to relevant signal offsets once the noise suression, an imortant art of the calorimeter signal extraction strategy resented in Sect. 5, is alied. Corrections mitigating the effect of these signal offsets on the reconstructed energy are discussed in the context of the ile-u suression strategy in Sect. 6.. All details of the ALAS liquid argon calorimeter readout and signal rocessing can be found in Ref. [9]. he ile calorimeter shows very little sensitivity to ileu since most of the associated (soft article) energy flow is absorbed in the LAr calorimeters in front of it. Moreover, out-of-time ile-u is suressed by a short shaing time with sensitivity to only about bunch crossings []. Monte Carlo simulation of s in the ALAS detector he energy and direction of articles roduced in roton roton collisions are simulated using various MC event generators. An overview of these generators for LHC hysics can be found in Ref. []. he samles using different event generators and theoretical models are described below. All samles are roduced at s = 7eV.. Inclusive Monte Carlo simulation samles. Pythia (version 6.45) [] is used for the generation of the baseline simulation event samles. It models the hard sub-rocess in the final states of the generated roton roton collisions using a matrix element at leading order in the strong couling α S. Additional radiation is modelled in the leading logarithmic (LL) aroximation by -ordered arton showers []. Multile arton interactions (MPI) [4], as well as fragmentation and hadronisation based on the Lund string model [5], are also generated. Relevant arameters for the modelling of the arton shower and multile arton interactions in the underlying event (UE) are tuned to LHC data (ALAS Pythia tune AUEB [6] with the MRS LO** arton density function (PDF) [7]). Data from the LEP collider are included in this tune.. Herwig++ [8] is used to generate samles for evaluating systematic uncertainties. his generator uses a matrix element and angular-ordered arton showers in the LL aroximation [9 ]. he cluster model [] is emloyed for the hadronisation. he underlying event and soft inclusive interactions are described using a hard and soft MPI model []. he arton densities are rovided by the MRS LO** PDF set.. MadGrah [4] with the CEQ6L PDF set [5]isused to generate roton roton collision samles with u to three outgoing artons from the matrix element and with MLM matching [6] alied in the arton shower, which is erformed with Pythia using the AUEB tune.. Z- and γ - Monte Carlo simulation samles. Pythia (version 6.45) is used to roduce Z- events with the modified leading-order PDF set MRS LO**. he simulation uses a matrix element to model the hard sub-rocess, and, as for the inclusive simulation, -ordered arton showers to model additional arton radiation in the LL aroximation. In addition, weights are alied to the first branching of the shower, so as to bring agreement with the matrix-element rate in the hard emission region. he same tune and PDF is used as for the inclusive samle.. he Algen generator (version.) [7]is used to roduce Z- events, interfaced to Herwig (version 6.5) [] for arton shower and fragmentation into articles, and to Jimmy (version 4.) [8] to model UE contributions using the ALAS AUE tune [9], here with the CEQ6L [5] leading-order PDF set. Algen is a leading-order matrix-element generator for hard multiarton rocesses ( n) in hadronic collisions. Parton showers are matched to the matrix element with the MLM matching scheme. he CEQ6L PDF set is emloyed.. he baseline γ - samle is roduced with Pythia (version 6.45). It generates non-diffractive events using a matrix element at leading order in α S to model the hard sub-rocess. Again, additional arton radiation is modelled by -ordered arton showers in the LL aroximation. he modelling of non-erturbative hysics effects arising in MPI, fragmentation, and hadronisation is based on the ALAS AUEB MRS LO** tune. 4. An alternative γ - event samle is generated with Herwig (version 6.5) and Jimmy using the ALAS AUE tune and the MRS LO** PDF. It is used to evaluate the systematic uncertainty due to hysics modelling. 5. he systematic uncertainty from s which are misidentified as hotons (fake hotons) is studied with a dedicated MC event samle. An inclusive samle is generated with Pythia (version 6.45) with the same arameter tuning and PDF set as the γ - samle. An additional filter is alied to the s built from the stable gener-

6 7 Page 6 of Eur. Phys. J. C (5) 75:7 ated articles to select events containing a narrow article, which is more likely to ass hoton identification criteria. he surviving events are assed through the same detector simulation software as the MC γ - samle.. o-quark air Monte Carlo simulation samles o air (t t) roduction samles are relevant for reconstruction erformance studies, as they are a significant source of hadronically decaying W bosons and theore imortant for light-quark resonse evaluations in a radiation environment very different from the inclusive and Z-/γ - samles discussed above. In addition, they rovide s from a heavy-flavour (b-quark) decay, the resonse to which can be studied in this final state as well. he nominal t t event samle is generated using MC@NLO (version 4.) [4], which imlements a nextto-leading-order (NLO) matrix element for to-air roduction. Corresondingly, the C [4] NLO PDF set is used. his matrix-element generator is interfaced to arton showers from Herwig (version 6.5) [4] and the underlying event modelled by Jimmy (version 4.), with the C PDF and the ALAS AUE tune. A number of systematic variation samles use alternative MC generators or different generator arameter sets. Additional t t samles are simulated using the POWHEG [4] generator interfaced with Pythia, as well as Herwig and Jimmy. POWHEG rovides alternative imlementations of the NLO matrix-element calculation and the interface to arton showers. hese samles allow comarison of two different arton shower, hadronisation and fragmentation models. In addition, the articular imlementations of the NLO matrix-element calculations in POWHEG and MC@NLO can be comared. Differences in the b-hadron decay tables between Pythia and Herwig are also significant enough to rovide a conservative uncertainty enveloe on the effects of the decay model. In addition, samles with more or less arton shower activity are generated with the leading-order generator ACERMC [44] interfaced to Pythia with the MRS LO** PDF set. hese are used to estimate the model deendence of the event selection. In these samles the initial state radiation (ISR) and the final state radiation (FSR) arameters are varied in value ranges not excluded by the current exerimental data, as detailed in Refs. [45,46]..4 Minimum bias samles Minimum bias events are generated using Pythia8[47] with the 4C tune [48] and MRS LO** PDF set. hese minimum bias events are used to form ile-u events, which are overlaid onto the hard-scatter events following a Poisson distribution around the average number μ of additional roton roton collisions er bunch crossing measured in the exeriment. he LHC bunch train structure with 6 roton bunches er train and 5 ns sacing between the bunches, is also modelled by organising the simulated collisions into four such trains. his allows the inclusion of out-of-time ile-u effects driven by the distance of the hard-scatter events from the beginning of the bunch train. he first ten bunch crossings in each LHC bunch train, aroximately, are characterised by varying out-of-time ile-u contributions from the collision history, which is getting filled with an increasing number of bunch crossings with roton roton interactions. For the remaining 6 bunch crossings in a train, the effect of the out-of-time ile-u contribution is stable, i.e. it does not vary with the bunch osition within the bunch train, if the bunchto-bunch intensity is constant. Bunch-to-bunch fluctuations in roton intensity at the LHC are not included in the simulation..5 Detector simulation he Geant4 software toolkit [49] within the ALAS simulation framework [5] roagates the stable articles roduced by the event generators through the ALAS detector and simulates their interactions with the detector material. Hadronic showers are simulated with the QGSP_BER model [5 59]. Comared to the simulation used in the context of the data analysis, a newer version of Geant4 (version 9.4) is used and a more detailed descrition of the geometry of the LAr calorimeter absorber structure is available. hese geometry changes introduce an increase in the calorimeter resonse to ions below GeV of about %. For the estimation of the systematic uncertainties arising from detector simulation, several samles are also roduced with the ALAS fast (arameterised) detector simulation ALFAS [5,6]. 4 Dataset he data used in this study were recorded by ALAS between May and October, with all ALAS subdetectors oerational. he corresonding total integrated luminosity is about 4.7 fb of roton roton collisions at a centre-of-mass energy of s = 7eV. As already indicated in Sect..4, the LHC oerated with bunch crossing intervals of 5 ns, and bunches organised in bunch trains. he average number of interactions er bunch See the discussion of truth s in Sect. 5.5 for the definition of stable articles.

7 Eur. Phys. J. C (5) 75:7 Page 7 of 7 otal Noise (MeV) 4 ALAS Simulation s = 7 ev, μ = PS EM EM EM ile ile ile Ga FCal FCal FCal HEC HEC HEC HEC (a) η otal Noise (MeV) 4 ALAS Simulation 5 ns bunch sacing s = 7 ev, μ = 8 PS EM EM EM ile ile ile Ga FCal FCal FCal HEC HEC HEC HEC (b) η Fig. he energy-equivalent cell noise in the ALAS calorimeters on the electromagnetic (EM) scale as a function of the direction η in the detector, for the configuration with a μ = and the configuration with b μ = 8. he various colours indicate the noise in the re-samler (PS) and the u to three layers of the LAr EM calorimeter, the u to three layers of the ile calorimeter, the four layers for the hadronic end-ca (HEC) calorimeter, and the three modules of the forward (FCal) calorimeter crossing (μ) as estimated from the luminosity measurement is μ 8 until Summer, with an average for this eriod of μ 6. Between August and the end of the roton run, μ increased to about 5 μ 7, with an average μ. he average number of interactions for the whole dataset is μ =8. he secific trigger requirements and recision signal object selections alied to the data are analysis deendent. hey are theore discussed in the context of each analysis resented in this aer. 5 Jet reconstruction and calibration with the ALAS detector 5. oological clusters in the calorimeter Clusters of energy deosits in the calorimeter (too-clusters) are built from toologically connected calorimeter cells that contain a significant signal above noise, see Refs. [,6,6] for details. he too-cluster formation follows cell signal significance atterns in the ALAS calorimeters. he signal significance is measured by the absolute ratio of the cell signal to the energy-equivalent noise in the cell. he signalto-noise thresholds for the cluster formation are not changed with resect to the settings given in Ref. []. However, the noise in the calorimeter increased due to the resence of multile roton-roton interactions, as discussed in Sect.., and required the adjustments exlained below. While in ALAS oerations rior to the cell noise was dominated by electronic noise, the short bunch crossing interval in LHC running added a noise comonent from bunch-to-bunch variations in the instantaneous luminosity and in the energy deosited in a given cell from revious collisions inside the window of sensitivity of the calorimeters. he cell noise thresholds steering the too-cluster formation thus needed to be increased from those used in to accommodate the corresonding fluctuations, which is done by raising the nominal noise according to σ noise = σnoise electronic (σ electronic noise ) + (σ ile-u noise ( oerations) ) ( oerations). Here, σnoise electronic is the electronic noise, and σ ile-u noise the noise from ile-u, determined with MC simulations and corresonding to an average of eight additional roton roton interactions er bunch crossing (μ = 8) in. he change of the total nominal noise σ noise and its deendence on the calorimeter region in ALAS can be seen by comaring Fig. a and b. In most calorimeter regions, the noise induced by ile-u is smaller than or of the same magnitude as the electronic noise, with the excetion of the forward calorimeters, where σ ile-u noise σnoise electronic. he imlicit noise suression imlemented by the toological cluster algorithm discussed above leads to significant imrovements in the calorimeter erformance for e.g. the energy and satial resolutions in the resence of ile-u. On the other hand, contributions from larger negative and ositive signal fluctuations introduced by ile-u can survive in a given event. hey thus contribute to the sensitivity to ileu observed in the resonse, in addition to the cell-level effects mentioned in Sect Jet reconstruction and calibration Jets are reconstructed using the anti-k t algorithm [] with distance arameters R =.4 or R =.6, utilising the FastJet software ackage [6,64]. he four-momentum scheme is used at each recombination ste in the clustering. he

8 7 Page 8 of Eur. Phys. J. C (5) 75:7 Jet reconstruction constituents s Simulated articles Jet finding ruth s racks Jet finding rack s Calorimeter clusters (EM scale) Local cluster weighting Jet finding Calorimeter clusters (LCW scale) Jet finding Calorimeter s (EM scale) Calorimeter s (LCW scale) Calibrates clusters based on cluster roerties related to shower develoment Fig. Overview of the ALAS reconstruction. After the finding, the four momentum is defined as the four momentum sum of its constituents Calorimeter s (EM or LCW scale) Pile-u offset correction Origin correction Energy & calibration Residual in situ calibration Calorimeter s (EM+JES or LCW+JES scale) Corrects for the energy offset introduced by ile-u. Deends on µ and NPV. Derived from MC. Changes the direction to oint to the rimary vertex. Does not affect the energy. Calibrates the energy and seudoraidity to the article scale. Derived from MC. Residual calibration derived using in situ measurements. Derived in data and MC. Alied only to data. Fig. Overview of the ALAS calibration scheme used for the dataset. he ile-u, absolute JES and the residual in situ corrections calibrate the scale of the, while the origin and the η corrections affect the direction of the total four-momentum is theore defined as the sum of the four-momenta sum of all its constituents. he inuts to the algorithm are stable simulated articles (truth s, see Sect. 5.5 for details), reconstructed tracks in the inner detector (track s, see Ref.[] and Sect. 5.4 for details) or energy deosits in the calorimeter (calorimeter s, see below for details). A schematic overview of the ALAS reconstruction is resented in Fig.. he calorimeter s are built from the too-clusters entering as massless articles in the algorithm as discussed in the revious section. Only clusters with ositive energy are considered. he too-clusters are initially reconstructed at the EM scale [6,65 7], which correctly measures the energy deosited in the calorimeter by articles roduced in electromagnetic showers. A second too-cluster collection is built by calibrating the calorimeter cell such that the resonse of the calorimeter to hadrons is correctly reconstructed. his calibration uses the local cell signal weighting (LCW) method that aims at an imroved resolution comared to the EM scale by correcting the signals from hadronic deosits, and thus reduces fluctuations due to the non-comensating nature of the ALAS calorimeter. he LCW method first classifies too-clusters as either electromagnetic or hadronic, rimarily based on the measured energy density and the longitudinal shower deth. Energy corrections are derived according to this classification from single charged and neutral ion MC simulations. Dedicated corrections address effects of calorimeter noncomensation, signal losses due to noise threshold effects, and energy lost in non-instrumented regions close to the cluster []. Figure shows an overview of the ALAS calibration scheme for calorimeter s used for the dataset, which restores the energy scale to that of s reconstructed from stable simulated articles (truth article level, see Sect. 5.5). his rocedure consists of four stes as described below.. Pile-u correction Jets formed from too-clusters at the EM or LCW scale are first calibrated by alying a correction to account for the energy offset caused by ile-u interactions. he effects of ile-u on the energy scale are caused by both additional roton collisions in a recorded event (intime ile-u) and by ast and future collisions influencing

9 Eur. Phys. J. C (5) 75:7 Page 9 of 7 Jet resonse at EM scale Barrel ALAS Simulation Barrel-endca ransition JES: Anti-k t R =.4, EM+JES HEC HEC-FCal ransition E = GeV E = 6 GeV E = GeV FCal E = 4 GeV E = GeV Jet η det (a) EMscale ( EM (η det )) Jet resonse at LCW scale Barrel ALAS Simulation Barrel-endca ransition JES: Anti-k t R =.4, LCW+JES HEC HEC-FCal ransition E = GeV E = 6 GeV E = GeV FCal E = 4 GeV E = GeV Jet η det (b) LCWscale ( LCW (η det )) Fig. 4 Average resonse of simulated s formed from too-clusters, calculated as defined in Eq. () and shown in a for the EM scale (R EM ) and in b for the LCW scale (R LCW ). he resonse is shown searately for various truth- energies as function of the uncorrected (detector) seudoraidity η det. Also indicated are the different calorimeter regions. he inverse of R EM (R LCW ) corresonds to the average energy scale correction for EM (LCW) in each η det bin. he results shown are based on the baseline Pythia inclusive samle the energy deosited in the current bunch-crossing (outof-time ile-u), and are outlined in Sect. 6. his correction is derived from MC simulations as a function of the number of reconstructed rimary vertices (N PV, measuring the actual collisions in a given event) and the exected average number of interactions (μ, sensitive to out-oftime ile-u) in bins of seudoraidity and transverse momentum (see Sect. 6).. Origin correction A correction to the calorimeter direction is alied that makes the ointing back to the rimary event vertex instead of the nominal centre of the ALAS detector.. Jet calibration based on MC simulations Following the strategy resented in Ref. [], the calibration of the energy and seudoraidity of a reconstructed is a simle correction derived from the relation of these quantities to the corresonding ones of the matching truth (see Sect. 5.5) in MC simulations. It can be alied to s formed from too-clusters at EM or at LCW scale with the resulting s being erred to as calibrated with the EM+JES or with the LCW+JES scheme. his first JES correction uses isolated s from an inclusive MC samle including ile-u events (the baseline samle described in Sect. ). Figure 4 shows the average energy resonse R EM(LCW) = E EM(LCW) /E truth, () which is the inverse of the energy calibration function, for various energies as a function of the seudoraidity η det measured in the detector frame of erence (see Sect. 5.6). 4. Residual in situ corrections A residual correction derived in situ is alied as a last ste to s reconstructed in data. he derivation of this correction is described in Sect Jet quality selection Jets with high transverse momenta roduced in roton roton collisions must be distinguished from background candidates not originating from hard-scattering events. A first strategy to select s from collisions and to suress background is resented in Ref. []. he main sources of otential background are:. Beam-gas events, where one roton of the beam collides with the residual gas within the beam ie.. Beam-halo events, for examle caused by interactions in the tertiary collimators in the beam-line far away from the ALAS detector.. Cosmic-ray muons overlaing in-time with collision events. 4. Calorimeter noise. he quality selection criteria should efficiently reject s from these background rocesses while maintaining high efficiency for selecting s roduced in roton roton collisions. Since the level and comosition of background deend on the event toology and the kinematics, four sets of criteria called Looser, Loose, Medium and ight are introduced in Ref. [7]. hey corresond to different levels of fake- rejection and selection efficiency, with the Looser criterion being the one with the highest selection efficiency while the ight criterion is the one with the best rejection. he discrimination between s coming from the collisions and background candidates is based on several ieces of exerimental information, including the quality of the energy reconstruction at the cell level, energy deosits in the direction of the shower develoment, and reconstructed tracks matched to the s. he efficiencies of the selection criteria are measured using the tag-and-robe method described in Ref. []. he

10 7 Page of Eur. Phys. J. C (5) 75:7 Jet selection efficiency ALAS Data, s = 7 ev η <. Looser Loose Medium ight Jet selection efficiency ALAS Data, s = 7 ev. η <.8 Looser Loose Medium ight Jet selection efficiency ALAS Data, s = 7 ev.8 η <. Looser Loose Medium ight data-mc (a) η <. data-mc (b). η <.8 (c).8 η <. data-mc Jet selection efficiency data-mc ALAS Data, s = 7 ev. η <. Looser Loose Medium ight Jet selection efficiency data-mc ALAS Data, s = 7 ev. η <.5 Looser Loose Medium ight (d). η <. (e). η <.5 (f).5 η <.8 Jet selection efficiency data-mc ALAS Data, s = 7 ev.5 η <.8 Looser Loose Medium ight Jet selection efficiency ALAS Data, s = 7 ev.8 η <.6 Looser Loose Medium ight Jet selection efficiency ALAS Data, s = 7 ev.6 η <4.5 Looser Loose Medium ight data-mc (g).8 η <.6 (h).6 η < 4.5 data-mc Fig. 5 Jet quality selection efficiency for anti-k t s with R =.4 measured with a tag-and-robe technique as a function of in various η ranges, for the four sets of selection criteria. Only statistical uncertainties are shown. Differences between data and MC simulations are also shown resulting efficiencies for anti-k t s with R =.4 for all selection criteria are shown in Fig. 5. he selection efficiency of the Looser selection is greater than 99.8 % over all calibrated transverse momenta and η bins. A slightly lower efficiency of about % is measured for the Loose selection, in articular at low and for.5 < η <.6. he Medium and ight selections have lower selection efficiencies mainly due to cuts on the charged fraction, which is the ratio of the scalar sum of the of all reconstructed tracks matching the, and the itself, see Ref. [7] for more details. For s with 5 GeV, the Medium and ight selections have inefficiencies of 4 and

11 Eur. Phys. J. C (5) 75:7 Page of 7 5 %, resectively. For > 5 GeV, the Medium and ight selections have efficiencies greater than 99 and 98 %, resectively. he event selection is based on the azimuthal distance between the robe and tag φ(tag, robe) and the significance of the missing transverse momentum E miss [74] reconstructed for the event, which is measured by the ratio E miss / E.Here E is the scalar transverse momentum sum of all articles, s, and soft signals in the event. he angle φ(tag, robe), E miss / E, and the ight selection of the erence (tag) are varied to study the systematic uncertainties. For the Looseand Looser selections, the selection efficiency is almost unchanged by varying the selection cuts, with variations of less than.5 %. Slightly larger changes are observed for the two other selections, but they are not larger than. % for the Medium and.5 % for the ight selection. he selection efficiency is also measured using a MC simulation samle. A very good agreement between data and simulation is observed for the Looser and Looseselections. Differences not larger than. and % are observed for the Medium and ight selections, resectively, for > 4 GeV. Larger differences are observed at lower,but they do not exceed % ( %) for the Medium(ight) selection. 5.4 rack s In addition to the reviously described calorimeter s reconstructed from too-clusters, track s in ALAS are built from reconstructed charged article tracks associated with the reconstructed rimary collision vertex, which is defined by ( track ) = max. Here track is the transverse momentum of tracks ointing to a given vertex. he tracks associated with the rimary vertex are required to have track > 5 MeV and to be within η <.5. Additional reconstruction quality criteria are alied, including the number of hits in the ixel detector (at least one) and in the silicon microstri detector (at least six) of the ALAS ID system. Further track selections are based on the transverse (d, erendicular to the beam axis) and longitudinal (z, along the beam axis) imact arameters of the tracks measured with resect to the rimary vertex ( d <.5 mm, z sin θ <.5 mm). Here θ is the olar angle of the track. Generally, track s used in the studies resented in this aer are reconstructed with the same configurations as calorimeter s, i.e. using the anti-k t algorithm with R =.4 and R =.6. As only tracks originating from the hardest rimary vertex in the collision event are used in the finding, the transverse momentum of any of these track s rovides a rather stable kinematic erence for matching calorimeter s, as it is indeendent of the ile-u activity. rack s can of course only be formed within the tracking detector coverage ( η <.5), yielding an effective accetance for track s of η track <.5 R. Certain studies may require slight modifications of the track selection and the track- formation criteria and algorithms. hose are indicated in the resective descritions of the alied methods. In articular, track s may be further selected by requirements concerning the number of clustered tracks, the track-, and the track- direction. 5.5 ruth s ruth s can be formed from stable articles generated in MC simulations. In general those are articles with a lifetime τ defined by cτ > mm [75]. he definitions alied are the same as the ones used for calorimeter and track s (anti-k t with distance arameters R =.4 and R =.6, resectively). If truth s are emloyed as a erence for calibrations uroses in MC simulations, neither final-state muons nor neutrinos are included in the stable articles considered for its formation. he simulated calorimeter s are calibrated with resect to truth s consisting of stable articles leaving an observable signal (visible energy) in the detector. 4 his is a articular useful strategy for inclusive measurements and the universal calibration discussed in this aer, but secial truth- erences including muons and/or neutrinos may be utilised as well, in articular to understand the heavy-flavour resonse, as discussed in detail in Sect Jet kinematics and directions Kinematic roerties of s relevant for their use in final-state selections and final-state reconstruction are the transverse momentum and the raidity y. he full reconstruction of the kinematics including these variables takes into account the hysics frame of erence, which in ALAS is defined event-by-event by the rimary collision vertex discussed in Sect On the other hand, many effects corrected by the various JES calibrations discussed in this aer are highly localised, i.e. they are due to secific detector features and inefficiencies at certain directions or ranges. he relevant directional variable to use as a basis for these corrections is then the detector 4 Muons can generate an observable signal in some of the ALAS calorimeters, but it is generally small and usually not roortional to the actual muon energy loss. heir contribution to the truth- energy, which can be large, is excluded to avoid biases and tails in the resonse function due to occasionally occurring high- muons in the MCsimulated calibration samles.

12 7 Page of Eur. Phys. J. C (5) 75:7 seudoraidity η det, which is reconstructed in the nominal detector frame of erence in ALAS, and is centred at the nominal collision vertex (x =, y =, z = ). Directional relations to s, and e.g. between the constituents of and its rincial axis, can then be measured either in the hysics or the detector erence frame, with the choice deending on the analysis. In the hysics erence frame ((y,φ) sace) the distance between any two objects is given by R = ( y) + ( φ), () where y is the raidity distance and φ is the azimuthal distance between them. he same distance measured in the detector frame of erence ((η, φ) sace) is calculated as R = ( η) + ( φ), () where η is the distance in seudoraidity between any two objects. In case of s and their constituents (too-clusters or tracks), η = η det is used. All clustering algorithms used in ALAS aly the hysics frame distance in Eq. () in their distance evaluations, as s are considered to be massive hysical objects, and the clustering is intended to follow energy flow atterns introduced by the hysics of arton showers, fragmentation, and hadronisation from a common (article) source. In this context too-clusters and reconstructed tracks are considered seudo-articles reresenting the true article flow within the limitations introduced by the resective detector accetances and resolutions. 6 Jet energy correction for ile-u interactions 6. Pile-u correction method he ile-u correction method alied to reconstructed s in ALAS is derived from MC simulations and validated with in situ and simulation based techniques. he aroach is to calculate the amount of transverse momentum generated by ile-u in a in MC simulation, and subtract this offset O from the reconstructed at any given signal scale (EM or LCW). At least to first order, ile-u contributions to the signal can be considered stochastic and diffuse with resect to the true signal. heore, both in-time and out-of-time ile-u are exected to deend only on the ast and resent ile-u activity, with linear relations between the amount of activity and the ile-u signal. 6. Princial ile-u correction strategy o characterise the in-time ile-u activity, the number of reconstructed rimary vertices (N PV ) is used. he ALAS tracking detector timing resolution allows the reconstruction of only in-time tracks and vertices, so that N PV rovides a good measure of the actual number of roton roton collisions in a recorded event. For the out-of-time ile-u activity, the average number of interactions er bunch crossing (μ) at the time of the recorded events rovides a good estimator. It is derived by averaging the actual number of interactions er bunch crossing over a rather large window t in time, which safely encomasses the time interval during which the ALAS calorimeter signal is sensitive to the activity in the collision history ( t 6 ns for the liquid-argon calorimeters). he observable μ can be reconstructed from the average luminosity L over this eriod t, the total inelastic roton roton cross section (σ inel = 7.5 mb[76]), the number of colliding bunches in LHC (N bunch ) and the LHC revolution frequency ( f LHC )(see Ref. [77] for details): L σ inel μ =. N bunch f LHC he MC-based calibration is derived for a given (erence) ile-u condition 5 (NPV,μ ) such that O(N PV = NPV,μ= μ ) =. As the amount of energy scattered into a by ile-u and the signal modification imosed by the ile-u history determine O, a general deendence on the distances from the erence oint is exected. From the nature of ile-u discussed earlier, the linear scaling of O in both N PV and μ rovides the ansatz for a correction, O(N PV,μ,η det ) = (N PV,μ,η det ) truth = ( ) (η det ) N PV NPV + N PV μ (η det) (μ μ ) ( ) ( = α(η det ) N PV NPV + β(η det ) μ μ ) (4) Here, (N PV,μ,η det ) is the reconstructed transverse momentum of the (without the JES correction described in Sect. 5. alied) in a given ile-u condition (N PV,μ) and at a given direction η det in the detector. he true transverse momentum of the ( truth ) is available from the generated article matching a reconstructed in MC simulations. he coefficients α(η det ) and β(η det ) deend on η det, as both in-time and out-of-time ile-u signal contributions manifest themselves differently in different calorimeter regions, according to the following influences:. he energy flow from collisions into that region.. he calorimeter granularity and occuancy after toocluster reconstruction, leading to different accetances at cluster level and different robabilities for multile article showers to overla in a single cluster. 5 he articular choice for a working oint, here (N PV = 4.9,μ = 5.4), is arbitrary and bears no consequence for the correction method and its uncertainty.