Calorimeter Calibration and Jet Energy Scale

1
Calorimeter Calibration and Jet Energy Scale

• Jean-Francois Arguin
• November 28th, 2005
• Physics 252B, UC Davis

2
Outline

• Quick remainder of calorimetry
• Calibration before the experiment starts test
  beam
• Calibration when the experiment is running
• Hardware calibration
• Collider data
• Measuring jets at high-energy colliders
• Example of a physics measurement top quark mass

3
Basics of Calorimetry

• Incident particle creates a shower inside
  material
• Shower can be either electromagnetic or hadronic
• Energy is deposited in material through
  ionization/excitation

4
Basics of Calorimetry II

• Basic principle of calorimetry
  deposited energy is proportional to incident
  energy
• Calorimeter calibration translate detector
  response to incident energy

• Great feature of showers for detector use length
  is proportional to logE

5
Electromagnetic showers

• Created by incident photon and electron
• electrons emit bremstrahlung
• photons undergo pair production
• Length of shower expressed in term of X0
• X0 depends on material

• 95 containment requires typically about 20X0

6
Hadronic showers

• Created by incident charged pion, kaon, proton,
  etc
• Typical composition
• 50 EM (e.g. )
• 25 Visible non-EM energy
• 25 invisible energy (nuclear break-ups)
• Requires longer containment (expressed in ?)

7
Calorimeter detectors

• Detector hardware must
• Favor shower development
• Collect deposited energy
• Can do both at the same time (e.g. BaBar/Belle
  crystal calorimeters)
• Or have calorimeters with alternating passive and
  sensitive material

• Example of electron shower with lead absorber

8
Sampling calorimetry (Ex. CDF)

• Scintillators (sensitive material) emit lights
  with passage of ionizing particles
• Collect light deposited in sensitive material
  using wavelength shifter (WLS)
• WLS ? photomultipliers that convert light into
  electric signal

9
CDF Calorimeters Segmentation
10
CDF Calorimeters
11
Construction

• Go on and built the thing after it is designed!
• Many institutions in the world participate

12
First calibration test beam

• Take one calorimeter wedge, send beam of
  particles with known energy
• Obtain correspondence detector response ? energy
  in GeV
• A few towers only submitted to test beam
• Set absolute scale for all towers
• Relative scale for other towers obtained later

13
How does the test beam works (Ex. plug
calorimeter)
Why Muons?

• Performed at Fermilab meson beam facilities
• Beams characteristics
• Various types for EM and HAD showers electrons,
  pions, muons
• Various energy 5-120 GeV (electrons), 5-220 GeV
  (pions)
• Beams can be contaminated ? bias the calibration
  constants
• E.g. use Cherenkov detector in front of
  calorimeter to identify proton contamination in
  pion beam

14
Calorimeter response linearity

• Extract calibration constant for many energy
  point
• Can test linearity of calorimeter
• Can add artificial material in front of
  calorimeter to simulate trackermagnet material
• Send pions and electrons to hadronic calorimeter

Why sending electrons in hadron calorimeter?
15
Performance determined from test beam

• From RMS of tower response to same beam energy ?
  measure calorimeter resolution
• Can test tower transverse uniformity (influences
  resolution)
• Stochastic term resolution
• EM
• HAD

16
Final detector assembly getting ready for
physics!
17
The Tevatron

• Proton-antiproton collisions at
• Most energetic collider in the world
• Collisions every 0.4 µs
• Circumference of 6.3 km

18
The CDF Detector

• CDF II general purpose
• solenoidal detector
• 7 layers of silicon tracking
• Vertexing, B-tagging
• COT drift chamber
• coverage
• Resolution
• Muon chambers
• Proportional chamber interspersed with absorber
• Provide muon ID up-to
• Calorimeters
• Central, wall, plug calorimeter

19
Calibration when the detector is installed

• Only a few towers saw test beam, how to calibrate
  the whole thing??
• Test beam sets the absolute scale as a function
  energy
• Two solutions
• Hardware calibrations
• Physics calibration (using collider data)
• These calibrations need to
• Cross-check absolute scale (e.g. test beam not
  100 realistic)
• Track detector response through time
• Expected degradation of scintillator and PMT
• PMT sensitive to temperature
• Uniform response through all towers

20
Hardware calibration

• Can use radioactive sources that have very well
  defined decay energy
• Cobalt 60 (2.8 MeV)
• Cesium 137 (1.2 MeV)
• Source calibration can be performed between
  colliders run
• Sources are movable and can expose one tower at a
  time
• Check uniformity over all towers and over time
• Sources are sensitive to both scintillator and
  PMT responses

21
Laser calibrations

• The lasers are connected directly to PMTs
• Skip scintillator/WLS steps
• Used to uniformize PMTs response over towers and
  time

22
Physics calibrations

• Use real collider data
• For calibration, you have to have some known
  and some unknown (the calorimeter response)
• Examples of known information
• Mass of a well-known particles
• Ex. Z?ee (Z mass measured at LEP)
• Energy deposited by muons over a given length
• Muon sample
• Energy measured in tracker (assuming tracker in
  calibrated)
• Redundant to energy measured in calorimeter for
  electrons

23
Example Z boson mass
Z mass peak

• Z mass measured with great accuracy at LEP using
  beam energy
• Background is very small for Z?ee
• Sample is relatively small, but good enough

24
Example E/p of electrons

• Used for relative scale over towers
• Cannot be used in forward region (no tracker)
• In plug rely on sourcing and lasers

25
Example muons for HAD calorimeters

• Muon calibrate detector response to ionizing
  energy
• Use muon from J/? for identification (mass not
  used like Z boson)
• Again, not used for PHA (rely on sourcing, laser)

26
Physics with photons/electrons
Search for new physics Z' candidate

• Calorimeter calibration not the only issue
• Electron/photon physics also rely on tracking
• Removal of background
• E.g. remove pion background by studying shower
  shape

Precision measurement W mass
27
What are jets?
Why not using tracker (has better resolution)?

• Jets are a collimated group of particles that
  result from the fragmentation of quarks and
  gluons
• They are measured as clusters in the calorimeter
• momentum of cluster of towers is correlated with
  the momentum of the original quark and lepton

28
Phenomenology of jets

• Quark/gluon produced from ppbar interaction
• Fragmentation into hadrons
• Jets clustering algorithm
• Adds towers inside cone
• Fraction of energy is out-of-cone
• Underlying event contributes

29
Jet versus calorimeter energy scale

• Jets are complicated processes
• Previous calorimeter calibrations are not
  sufficient to get calibrated jet energy
• More work needs to be done!!
• Jet energy scale is crucial for many important
  measurements
• Top quark mass (used to constrain Higgs boson)
• Jet cross-sections (comparison to QCD
  predictions)
• Measurements often performed by comparing real
  data with simulations
• Need to get both physics and detector simulation
  right

30
Relative energy scale

• Relative energy scale
• Use QCD dijet events
• Should have equal transverse momentum

• Jet energy measurement depend on location in
  detector
• True even after all previous calibrations!
• How come?
• Jets are wide
• Some regions of CDF calorimeter are not
  instrumented

31
Absolute energy scale

• Solution
• Get the average energy scale Simulate an
  average particles configuration inside jet
• Use test beam information to get calibration
  factor for single particles

• Response to single pion non-linear (in test beam)
• However, jets are identified as one single
  objects
• For a 50 GeV jet calibration is not the same
  whether
• One 50 GeV pion
• 10 times 5 GeV pions

32
Out-of-cone energy

• Cone of fixed radius used to identify jets
• Need to correct for fraction of energy
  out-of-cone (typically 15)
• This is mostly physics related
• How well is the physics generator representing
  fragmentation?

33
Underlying event energy

• Proton/antiproton remnants splash energy in
  calorimeter
• Spoils jet energy measurement
• Depends on the number of ppbar interaction per
  event
• Extracted from minimum bias events
• Small effect 0.4 GeV per jet

34
Final jet energy scale uncertainty

• Estimate of jet energy scale uncertainty is
  important to estimate systematic uncertainties of
  measurements
• Dominated by out-of-cone (low-pT) and absolute
  energy scale (high-pT)
• Ranges from 10 to 3 energy uncertainties

35
Example physics measurement top quark mass

• Top produced in pairs at Tevatron
• Top decays to W boson and b-quark 100 of time in
  SM
• Typical event selections
• Well-identified electron(s) or muon(s)
• Large missing ET
• Several reconstructed jets identified in
  calorimeters
• Note 4 jets in final state!

36
Identification of b-quark jets

• Complicated final state
• Which jets come from which parton?
• Can identify b-quark jets using one
  characteristic
• Long b-quark lifetime
• Note lots of semileptonc B-hadrons decay
  (involving neutrino)
• Require special b-jets calibration

37
Top mass reconstruction

• Event-by-event kinematic fitter (assumes event is
  ttbar)
• Attempts all jet-parton assignments
• Assign b-tag jets to b-quarks
• The one most consistent with ttbar hypothesis is
  kept

More correct combinations with b-tags!
38
The strategy

• Construct reconstructed top mass distributions
  for many true top mass
• So-called templates
• Compare distribution reconstructed in data with
  templates
• Using likelihood fit
• Account for background contamination
• Dominated by Wjets production

39
The measurement (spring 2005)

• Using 138 candidate ttbar events, fit yields
• Mtop 173.2 2.9/-2.8 (stat.) /- 3.4
  (syst.) GeV/c2

• By shifting by JES uncertainty defined before
  Mtop changes by 3.1 GeV/c !
• JES uncertainty limiting factor for Mtop
  measurement

40
Improvement W?jj calibration

• Inside ttbar events, invariant mass of two jets
  from W boson decay should equal MW
• Can use W?jj decays to further constraint JES
• Use same data for measurement and calibration
  cheating??
• No Mjj (almost) independent of Mtop
• Remaining correlations are accounted for

41
The measurement(adding W?jj information)

• Using same dataset as previously
• Mtop 173.5 2.7/-2.6 (stat.) /- 2.8
  (syst.) GeV/c2

• Total Mtop uncertainty improved by 10
• JES uncertainty decreased by 20
• Good prospect for future

42
Impact of Mtop measurement

• Mtop, MW connected to Higgs boson mass through
  radiative corrections
• MHlt 186 GeV/c2 _at_ 95C.L.
• Can constrain mass of supersymmetric particles

43
Conclusion

• Detector calibration needed to translate detector
  response in energy
• Various techniques used for calorimetry
• Test beam
• Radioactive sources
• Lasers
• Collider data
• Calorimeter can be used to measure
• Electrons, photons, jets, missing ET
• Good calorimeter and jet calibration needed for
  measurements like top quark mass