| A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | JEC Name | Short Description from Run2 twiki | Correlated CMS-ATLAS (Run1)? | Why / Motivation (Run1 CMS-ATLAS) from JME-15-001 | Correlated CMS-ATLAS (Run2)? | Why / Motivation | ||||||||||||||||
2 | AbsoluteStat | Statistical uncertainty on in situ calibration with photon-jet/Z-jet events | 0% (uncorrelated) - statistical in situ term | Group 1 'Statistical and detector-based uncertainties from in situ techniques': ATLAS and CMS use statistically distinct data and MC samples and different detector technologies, so these are uncorrelated. | 0% (uncorrelated) - statistical in situ term-> still for Run2 | KEEP SAME MOTIVATION AS: JME-15-001Group 1 'Statistical and detector-based uncertainties from in situ techniques': ATLAS and CMS use statistically distinct data and MC samples and different detector technologies, so these are uncorrelated. | ||||||||||||||||
3 | AbsoluteScale | Flat absolute scale uncertainty (combined photon, Z->ee, Z->mumu reference scale) | 0% (uncorrelated) - detector in situ term | Group 1 (detector-based, Table 4 row 1b): detector-description-driven terms are treated the same as the statistical ones - fixed at 0% since they rely on each experiment's own detector calibration. | 0% (uncorrelated) - detector in situ term -> still for Run2 | KEEP SAME MOTIVATION AS: JME-15-001Group 1 (detector-based, Table 4 row 1b): detector-description-driven terms are treated the same as the statistical ones - fixed at 0% since they rely on each experiment's own detector calibration. | ||||||||||||||||
4 | AbsoluteMPFBias | MPF method bias correction (neutrinos and ISR+FSR outside detector acceptance) | 0-50% (partially correlated) - absolute balance modeling term | Group 2 'Modeling uncertainties from in situ techniques': ATLAS and CMS measure similar effects (neutrinos in jets, out-of-cone, radiation) but combine them differently (ATLAS: pT-binned combination; CMS: global 2-parameter fit) -> partially correlated, scanned 0-50%. | 0-50% (partially correlated) - absolute balance modeling term -> still for Run2 | KEEP SAME MOTIVATION AS: JME-15-001 Group 2 'Modeling uncertainties from in situ techniques': ATLAS and CMS measure similar effects (neutrinos in jets, out-of-cone, radiation) but combine them differently (ATLAS: pT-binned combination; CMS: global 2-parameter fit) -> partially correlated, scanned 0-50%. | ||||||||||||||||
5 | Fragmentation | High-pT extrapolation based on Pythia6 Z2 / Herwig++2.3 differences in fragmentation and underlying event (FullSim). Which models have been used? Full uncorellated can be conservative | 0% (uncorrelated) - grouped with High-pT uncertainties. | Group 8 'High-pT uncertainties': the CMS Fragmentation term is mapped (Table 4, row 8) to ATLAS' High-pT uncertainty; methodology and test-beam energies used differ between experiments, so treated as uncorrelated. | ? | To be discussed with ATLAS. Did the ATLAS method change? Which tune they use (e.g. Pythia8)? | ||||||||||||||||
6 | SinglePionECAL | High-pT extrapolation based on propagation of ±3% variation in single particle response in ECAL to PF jets (FastSim) | 0% (uncorrelated) - detector in situ term | Group 1b 'Detector in situ terms': single-hadron response measurements are experiment-specific (different test-beam/detector data), fixed at 0%. | 0% (uncorrelated) - detector in situ term -> still for Run2 | KEEP SAME MOTIVATION AS: JME-15-001 Group 1b 'Detector in situ terms': single-hadron response measurements are experiment-specific (different test-beam/detector data), fixed at 0%. | ||||||||||||||||
7 | SinglePionHCAL | High-pT extrapolation based on propagation of ±3% variation in single particle response in HCAL to PF jets (FastSim) | 0% (uncorrelated) - statistical in situ term | Group 1a 'Statistical in situ terms': in Run1 this component is constrained by the CMS global fit and becomes a statistical term, so grouped with the purely statistical/detector uncertainties, fixed at 0%. | 0% (uncorrelated) - statistical in situ term -> still for Run2 | KEEP SAME MOTIVATION AS: JME-15-001 Group 1a 'Statistical in situ terms': in Run1 this component is constrained by the CMS global fit and becomes a statistical term, so grouped with the purely statistical/detector uncertainties, fixed at 0%. | ||||||||||||||||
8 | FlavorQCD | Default jet-flavor uncertainty, based on Pythia6 Z2/Herwig++2.3 differences in uds/c/b-quark and gluon responses | Not a single value - Run1 note splits flavor by parton type (see FlavorPureGluon/Quark/Charm/Bottom below). For Run2: we should consider the correlation? We need info from ATLAS for Run2 | Section 4 groups 4-6: the note does not treat 'flavor' as one component; it separates gluon-jet (group 4), b-jet (group 5) and other quark/charm flavors (group 6), each with its own correlation, because the modeling robustness and the ATLAS/CMS derivation methods differ by parton type. | ? | To be discussed with ATLAS. Is also ATLAS considering a gloabal source of uncertainty for FlavourQCD? The low response observed in Run 1 is expected to have a common component between CMS and ATLAS, while the underlying event has a relatively small impact. | ||||||||||||||||
9 | FlavorPureGluon | Alternative flavor uncertainty source, gluon-only jets | 100% (fully correlated) | Group 4 'g-jet fragmentation energy scale uncertainty': both experiments derive the uncertainty from the same PYTHIA/HERWIG++ difference; Figure 1 in the note shows the resulting ATLAS and CMS uncertainties are nearly identical across pT and |eta|<3.2, so the components are taken as fully correlated. | 100% (fully correlated)->still in Run2 | KEEP SAME MOTIVATION AS: JME-15-001 Group 4 'g-jet fragmentation energy scale uncertainty': both experiments derive the uncertainty from the same PYTHIA/HERWIG++ difference; Figure 1 in the note shows the resulting ATLAS and CMS uncertainties are nearly identical across pT and |eta|<3.2, so the components are taken as fully correlated. | ||||||||||||||||
10 | FlavorPureQuark | Alternative flavor uncertainty source, uds-only jets | 0% (uncorrelated) - other fragmentation types | Group 6 'Other jet fragmentation energy scale uncertainties': ATLAS folds light-quark differences into its larger gluon-uncertainty budget, while CMS derives a dedicated per-flavor PYTHIA/HERWIG++ comparison; the note states the two techniques are 'unrelated... and not directly comparable', hence uncorrelated. | 0% (uncorrelated) - other fragmentation types->still in Run2 | KEEP SAME MOTIVATION AS: JME-15-001 Group 6 'Other jet fragmentation energy scale uncertainties': ATLAS folds light-quark differences into its larger gluon-uncertainty budget, while CMS derives a dedicated per-flavor PYTHIA/HERWIG++ comparison; the note states the two techniques are 'unrelated... and not directly comparable', hence uncorrelated. | ||||||||||||||||
11 | FlavorPureCharm | Alternative flavor uncertainty source, charm-only jets | 0% (uncorrelated) - other fragmentation types | Group 6, same reasoning as FlavorPureQuark: ATLAS and CMS derivation methods for non-gluon, non-b flavors are not directly comparable, so treated as uncorrelated. | 0% (uncorrelated) - other fragmentation types -> still in Run2 | KEEP SAME MOTIVATION AS: JME-15-001 Group 6, same reasoning as FlavorPureQuark: ATLAS and CMS derivation methods for non-gluon, non-b flavors are not directly comparable, so treated as uncorrelated. | ||||||||||||||||
12 | FlavorPureBottom | Alternative flavor uncertainty source, bottom-only jets | 50-100% (mostly correlated) | Group 5 'b-jet fragmentation energy scale uncertainty': both derive the uncertainty from PYTHIA/HERWIG++ differences with substantial methodological similarity, but ATLAS parametrizes it vs pT (flat in eta) while CMS does the opposite, adding some decorrelation - hence mostly correlated, scanned 50-100%. | 50-100% (mostly correlated) -> still in Run2 | KEEP SAME MOTIVATION AS: JME-15-001 Group 5 'b-jet fragmentation energy scale uncertainty': both derive the uncertainty from PYTHIA/HERWIG++ differences with substantial methodological similarity, but ATLAS parametrizes it vs pT (flat in eta) while CMS does the opposite, adding some decorrelation - hence mostly correlated, scanned 50-100%. | ||||||||||||||||
13 | TimePtEta | JEC time dependence between data-taking epochs (BCD, EF, G, H); L2L3Residual now corrected per epoch. Full 2016 sample, the uncertainty is not necessary. | 0% (uncorrelated) - single-experiment term | Group 9 'Other uncertainties': a small number of remaining sources evaluated in one experiment but not the other are taken as uncorrelated between ATLAS and CMS; Table 4 explicitly maps TimeEta/TimePt here. Does atlas have this uncertainty? | 0% (uncorrelated) - single-experiment term-> still in Run2 | KEEP SAME MOTIVATION AS: JME-15-001 Group 9 'Other uncertainties': a small number of remaining sources evaluated in one experiment but not the other are taken as uncorrelated between ATLAS and CMS; Table 4 explicitly maps TimeEta/TimePt here. Does atlas have this uncertainty? | ||||||||||||||||
14 | RelativeJER[EC1][EC2][HF] | Eta-dependence uncertainty from jet pT resolution (JER), derived by varying JER scale factors up/down | 0% (uncorrelated) - detector in situ term -> still in Run2. ATLAS is using the same method of Run1? | Group 1b 'Detector in situ terms' (Table 4): JER data/MC differences depend on each experiment's own detector resolution measurement, fixed at 0%. | 0% (uncorrelated) - detector in situ term->still in Run2. To be checked | To be discussed with ATLAS. Does ATLAS use the same method of Run1 (to be checked)? | ||||||||||||||||
15 | RelativePt[BB][EC1][EC2][HF] | Half-difference between MPF log-linear (default) and constant fits vs pT | 0% (uncorrelated) - detector in situ term->still in Run2. What atlas uses for the parametrization? | Group 1b 'Detector in situ terms' (Table 4): treated alongside AbsoluteScale/SinglePionECAL/RelativeJER as a detector-driven, experiment-specific term, fixed at 0%. | 0% (uncorrelated) - detector in situ term->still in Run2. To be checked | To be discussed with ATLAS. What ATLAS used for the parametrisation uncertainty? | ||||||||||||||||
16 | RelativeBal | Full difference between log-linear fits of the dijet MPF and pT balance methods | N/A - source does not exist in the Run1 note, ATLAS introduces something equivalent? | N/A - RelativeBal is not defined anywhere in this note (Tables 3/4 or Section 3); it was introduced only for Run2, so no Run1 correlation can be quoted from this source. | ? | To be discussed with ATLAS. Did ATLAS introduce an equivalent uncertainty? | ||||||||||||||||
17 | RelativeSample | Eta-dependent uncertainty due to the difference between relative residuals observed with dijet, Z+jets and gamma+jets samples | N/A - source does not exist in the Run1 note, ATLAS introduces something equivalent? | N/A - same as RelativeBal: not present in this note's Table 3/4, introduced only in Run2. | ? | To be discussed with ATLAS. Did ATLAS introduce an equivalent uncertainty? | ||||||||||||||||
18 | RelativeFSR | Eta-dependence uncertainty due to correction for initial and final state radiation | 50-100% (mostly correlated) | Group 3 'Relative eta-intercalibration modeling uncertainty': both experiments use a generator-difference (PYTHIA vs HERWIG++ for CMS; similar approach for ATLAS) to model radiation in the dijet eta-intercalibration; given the similar generators/techniques, treated as mostly correlated, scanned 50-100%. | Decorellate? or partly correlate (0-50%)? | To be discussed with ATLAS. | ||||||||||||||||
19 | RelativeStat[FSR][EC][HF] | Statistical uncertainty in the determination of eta-dependence (FSR correction fit / L2Res log-linear fit) | 0% (uncorrelated) - statistical in situ term- -> still in Run2 | Group 1a 'Statistical in situ terms' (Table 4): purely statistical uncertainties from distinct data samples/detectors, fixed at 0%. | 0% (uncorrelated) - statistical in situ term- -> still in Run2 | KEEP SAME MOTIVATION AS: JME-15-001 Group 1a 'Statistical in situ terms' (Table 4): purely statistical uncertainties from distinct data samples/detectors, fixed at 0%. | ||||||||||||||||
20 | PileUpDataMC | 5% uncertainty on the data/MC scale factor for offset correction (roughly covers mu variation seen with Random Cone on Zero Bias data / neutrino gun MC) | 0% (uncorrelated) - pileup term->still in Run2 | Group 7 'Pileup uncertainties': although both use variants of FastJet area-based pileup subtraction, the actual derivation of the uncertainty differs dramatically (CMS: pT-dependent offset difference in/out of jets; ATLAS: 4 separate NPV/mu/rho-based terms), so treated as uncorrelated. | 0% (uncorrelated) - pileup term->still in Run2 | KEEP SAME MOTIVATION AS: JME-15-001 Group 7 'Pileup uncertainties': although both use variants of FastJet area-based pileup subtraction, the actual derivation of the uncertainty differs dramatically (CMS: pT-dependent offset difference in/out of jets; ATLAS: 4 separate NPV/mu/rho-based terms), so treated as uncorrelated. | ||||||||||||||||
21 | PileUpPt[Ref][BB][EC1][EC2][HF] | Pile-up offset dependence on jet pT, estimated from matched MC with/without PU overlay vs Random Cone method | 0% (uncorrelated) - pileup term->still in Run2 | Group 7, same reasoning as PileUpDataMC: pileup suppression is conceptually similar but the uncertainty-derivation methods differ substantially between the two experiments, so uncorrelated. | 0% (uncorrelated) - pileup term->still in Run2 | KEEP SAME MOTIVATION AS: JME-15-001 Group 7, same reasoning as PileUpDataMC: pileup suppression is conceptually similar but the uncertainty-derivation methods differ substantially between the two experiments, so uncorrelated. | ||||||||||||||||
22 | JER | N/A in the note | ? | To be discussed with ATLAS. | ||||||||||||||||||
23 | ||||||||||||||||||||||
24 | ||||||||||||||||||||||
25 | ||||||||||||||||||||||
26 | ||||||||||||||||||||||
27 | ||||||||||||||||||||||
28 |