Some rather general thoughts on WZlepton triggering and precision measurements' Terry Wyatt' Manches

1
Some rather general thoughts on W/Z?lepton
triggering and precision measurements.Terry
Wyatt. Manchester/FNAL.W/Z Meeting 24th May
2002

• Motivation to achieve per mille level systematic
  errors
• Redundancy
• Some vague examples (guesses) to provoke real
  work.

2
A physics measurement to provide the motivation
to do as well as we possibly can

• Measurement of Br(W?e?)/Br(W???)
• Direct test of lepton universality in W decays
  or
• Direct search for lepton non-universality in W
  decays
• Statistical precision 10-3
• Limited by systematic errors on trigger
  efficiency, event selection, background,
  bookkeeping and other biases
• An interesting physics measurement direct test
  of how well we understand systematic biases in
  other (even more interesting) measurements, e.g.,
  mW, mt.

3
General Features of W?e? Events

• EM calorimeter cluster
• high ET
• longitudinal and transverse
• shower shape
• preshower cluster
• energy
• shape
• quality of match to EM
• cluster

• central track
• pT
• quality of match
• to preshower
• and EM clusters

jet

• missing ET
• magnitude

• Plus
• isolation of electron and missing ET in EM cal,
  preshower and tracker
• mT(e?)

4
Redundancy - We dont need to require all of
these features to trigger or select every
candidate event

• A trivial but instructive example
• We have two uncorrelated methods of triggering
  events with efficiency 90?5 and 50?10
• Taking the .OR. gives efficiency 95?2.7
• Higher efficiency and smaller error
• We can measure the efficiency
• Ideal Aim There should be no specific feature
  that we absolutely rely on to select/trigger our
  events.

5
A slightly more real example W?e? triggering in
Level 3

• N.B. concentrate on the general idea rather than
• precise details (which are complete guesses).
• Any one of the following should trigger
  standalone
• ETem gt 30 GeV (either cone or road method)
• pTtrack gt 30 GeV
• missing ET gt 30 GeV

6

• ETem gt 25 GeV (either cone or road method) should
  trigger if accompanied by any one of
• e.m. cluster passes shower shape cuts
• e.m. cluster passes isolation cuts
• matched pTtrack gt 25 GeV
• missing ET gt 25 GeV
• matched preshower cluster
• ETem gt 20 GeV (either cone or road method) should
  trigger if accompanied by any two of the above
  (with 25 replaced with 20).
• ETem gt 15 GeV (either cone or road method) should
  trigger if accompanied by any three of the above
  (with 25 replaced with 15).

7
Obvious questions/objections

• This is all very complicated.
• This will place very high demands on
• detector and trigger performance and stability
• rigorous operating procedures and monitoring
  during datataking
• Doing the analysis will require
• a detailed knowledge of how the trigger works
• careful investigation of correlations and
  biasses
• Yes! Guilty on all counts. This is the challenge
  in doing per mille level measurements.
• (But of course, without the redundancy the
  measurement is even more sensitive to screw ups
  in any particular area.)

8
Why worry about this stuff now, when we are still
commissioning the basic detector/trigger systems?

• Trigger has limited bandwidth at L1/L2/L3.
• More redundancy leads potentially to higher
  rates.
• Trigger bandwidth allocated to high pT electrons
  and muons (and taus) is not available to QCD,
  B-physics, etc.
• We shall need quantitative studies to back up our
  trigger requests.

9
Particular area of concern

• Single muon triggers at L1, L2 and L3
• pT cut needed to control rates
• Poor local muon pT resolution could lead to
  over-reliance on tracking
• Need standalone muon system trigger at L1, L2 and
  L3
• At L1 beyond 8-layer CFT acceptance there is no
  redundancy at all!