HLT1 luminosity counters

We need to cover as many detectors as possible with luminosity counters in HLT 1. The counters available in HLT 1 have the benefit of being calculated during physics data taking, absolute calibrations (VDM/BGI) and emittance scans. Any final luminosity calibration would either use these counters directly or (worst case) use them to cross-calibrate other counters not available in HLT 1.

Some general guidelines for luminosity counters are:

• Independent is better then dependent
  • e.g. number of muon candidates without the tracking efficiency folded in
  • dependent counters still useful: they are not 100% correlated (e.g. T tracks depend only on the FT but might suffer from noise, whereas long tracks are clean but fold in VP and UT)
• High-level is better than low-level
  • noise pedestals (e.g. in hit distributions) are hard to deal with
  • better: objects corresponding to the physics interaction (track, PV, Z->mumu, ...)

A preliminary list of counters to implement (in order of importance)

• number of VELO tracks and PVs (with and w/o fiducial cuts)
• PLUME-based counters (see #196 (closed))
• Calorimeter-based counters (e.g. similar to L0 HadronEt)
• Tracker-based (e.g. T tracks, VeloUT tracks, long tracks, ...)
• Muon-based
• ...