Missing Et in rel17

The recommended missing Et is the MET_RefFinal as given by default in the rel17. It is calculated from calorimeter cells with abs(eta) < 4.9 and from muons. Cells are calibrated according to the object to which they are associated.

The objects are:

• Medium Electrons using the default calibration from the egamma group and pT> 10 GeV
• Tight photons calibrated at the EM scale and pT>10 GeV
• Tight taus with LCW calibration (for tau-based analyses: tau identification with BDTMedium with El_BDT_medium veto and muon veto, LCW calibration and TES applied).
• LC topo anti-kT R=0.4 jets (with pT > 20 GeV) using the default JES calibration of the Jet/Etmiss group
• LC topo anti-kT R=0.4 jets (with 10 < pT < 20 GeV), calibrated with LCW (not applying the JES)
• CellOut calibrated with LCW and using the track-cluster matching algorithm (aka CellOut_eflow) in the above-mentioned objects!
• combined muons are Staco muons, tagged muons are used not only in the transition region, better quality criteria for standalone MuonBoy muons in 2.5<|eta|<2.7

A more detailed description and configuration of all the METRef terms is available in the EtMissRefFinal Twiki A detailed description of the algorithm can be found in the ETmiss performance paper to be published in EPJC.

This MET_RefFinal had been validated in the full pseudo-rapidity range with Z, W, di-jet and Minimum bias events in rel16 2011 data and MC10b simulation. Further validation has been performed on Z events with rel 17 2011 data and MC11a/b simulation. All tests have proved that its performance is superior or equivalent with MET_LocHadTopo previously used in physics analyses.

The D3PDs produced from rel17 AODs have to contain:

• the object collections corresponding to the MET_REfFinal terms, e.g. the AntiKt4LCTopoJets used in the RefJet term.

Systematic uncertainties

The systematic uncertainties for MET_RefFinal have to be estimated by propagating the participating object uncertainties to the event MET_RefFinal.

• The code for this is available in the MissingETUtility package. Details can be found in the slides from TJ Khoo and this Twiki.
  • The uncertainties for electrons, muons, jets, taus are used as given by the relevant CP groups
  • The uncertainties on CellOut and SoftJets as a function of SumEt are built in the tool.
  • Additional uncertainty due to (for example) pileup can be also set. A 10% shift in terms is the default, but another value can be set.
• In order to use the uncertainties propagation tools on a D3PD, the D3PD has to contain:
  • All the MET_RefFinal terms
  • the METComposition map weights for the participating objects
  • for rel17 D3PDs remember to include the branches jet_AntiKt4LCTopo_MET_w*
• For use on AODs an Athena implementation in the form of an AthAlgTool is provided. See METUtilityAthena for details.

The MissingETUtility package also contains tools for scaling and smearing the electron/muon energies/momenta such that MET_RefFinal can be re-calculated with smeared/scaled such objects.

Note: In order to change the jet collection in METComposition for D3PD production, the following lines are needed:
from MissingETD3PDMaker.MissingETD3PDMakerFlags import *
MissingETD3PDMakerFlags.METDefaultJetCollectionSGKey = 'AntiKt4LCTopoJets'
MissingETD3PDMakerFlags.METDefaultJetPrefix = "jet_antikt4LCtopo_MET_"

Improvements and implementations in rel 17.2

* Implemented new MET_RefFinal based on third muon chain

* Improvements in MET_Track calculation, update the code to calculate MET_Track from tracks from the first PV vertex

* Improvements in AOD occupancy of MET objects, removed some obsolete objects from AODs, plus the composition map for the MET_RefFinal_em case. In any case all objects are available on ESDs.

* Change in electron identification in MissingET-03-03-29: moved to passID function and ElectronIDMediumPP

* Known problem in RefGamma that is always zero in MissingET-03-03-29 affecting T0 ESDs and AODs: problem in RefGamma at T0 Release.

We recommend to use tag MissingET-03-03-31-08 to reconstruct Etmiss from the AOD and to redo the D3PD.

Pileup suppression methods

-- IreneVichou - 09-Nov-2011