emid: p11 certification and beyond

1
em-id p11 certification and beyond
2
data samples for certification

• Data samples
• reconstructed with p11.11, p11.12
• calorimeter 0-suppression online 1.5?, offline
  2.5? (2.1?)
• removed runs with event mixing from 26thJune
  until 15th August
• encourage use of em-samples skimmed or picked
• by the W/Z and top group
• MC samples
• reconstructed with p10.11, p11.10 with1.5? or
  2.5? noise cut and p11.12
• but noise simulation in the calorimeter is
  underestimated by ?2-3
• urgent to have MC available with improved noise
  simulation and consistent 0-suppression cuts
• H-matrix determination, efficiency calculations,
  Data/MC comparisons

3
em-candidate changes
Volker Buescher, Mainz

• major changes since certification 2.1
• geometry dependent energy corrections
• inter-cryostat corrections added
• H-matrix
• obtained from p11.12 MC, but will not be
  certified!
• alignment corrections
• EC alignment, CPS alignment
• track-match in EC
• CPS-match

4
L1 trigger efficiencies

• emEffTool provides em-trigger efficiencies for
  any specified run range
• computes efficiencies w.r.t. unbiased ?-triggers
• matches em trigger-tower with good electrons (dR
  lt0.5)
• provides plateau value, root-file with
  histograms and macro to visualize the plots

Tobias Golling, Bonn
5
L2 em requirements for TL9.0

• reduce CEM(1,10) rate ? L2 ET threshold at 12
  GeV
• EMFR cut only small rejection obtained
• rejection w/ no EMFR cut is 1.8
• rejection w/ EMFRgt0.8 is 2.1
• calibration issues?
• rejection vs. threshold worse rejection with
  recent data!

Sean Mattingly, Brown

• trigger extension towards EC-region?

6
em-selection efficiency
Alex Melnitchouk, Brown

• determined from Z events with 1 tight electron
• estimate number of signal/background events in
  peak

Black whole sample Blue at least one jet
(JCCA) in the event
emfrgt0.9 isogt0.15 hmatrix(hmcut)
CC
EC
recompute H-matrix a.s.a. MC with improved
calorimeter noise available!
7
em-fake probability
Alex Melnitchouk, Brown
determined from ejets data sample
high fake prob. for ? how to improve?
8
em-estimators Data/MC
purpose study effect of 1.5/2.5 sigma noise cut
on em estimators problem underestimation of
noise in MC
Tuan Vu Ahn, Orsay
CC em-fraction
EC em-fraction

• Data
• 2.5 ?
• - 1.5 ?

9
em energy deposition
Tuan Vu Ahn, Orsay
fraction of energy deposited in 3rd layer
number of cells in 3rd layer
CC
CC
EC
EC

• non consistent description of energy deposited
  in cells
• fix noise simulation first other discrepancies
  can hide behind!

10
track-calorimeter match
Robert Zitoun, SUNY
determine resolution of track/cluster
match compute ?2 probability - cut P(?2) gt 1
11
track-calo match EC
Robert Zitoun, SUNY

• first track-match in EC E/p small contribution

12
track matching efficiency

• Using 125 Z events with 0, 1 and 2 matched tracks
• Efficiency vs. h tag Z with onematched cluster
  and look if oppositecluster is matched
  (sensitive tocorrelation)

Robert Zitoun, SUNY
13
CPS matching

• matching criteria determined similar to
  track-match
• cps-track match
• in a box 0.1 radian by 15cm find the highest
  energy cps cluster
• sDp 3.8 mrad
• sDz 3.5 mm
• performed same exercise from cal only info

cps-track ??
cps-track ?z
Drew Alton, Michigan
14
CPS Rejection/Efficiency
Drew Alton, Michigan

• ptgt15, IDlt 20, isolt0.15, Emfr gt0.9, HMx lt100
• Rejection
• Rungt151831 and 1 of two EM in CPS fiducial.
• 1 of two EM has cps_cal_match_probgt1e-4
• Each EM in CPS fiducial has match.
• Efficiency
• Track match
• Track and cps match
• Rejection (upper limit) 0.8(0.6)
• Efficiency 0.9

15
CPS alignment

• cps-matching possible after alignment
  determination in p11
• to come in p13
• improved geometry in simulation and
  reconstruction
• deghosting
• 3d fps clusters
• from detector side
• low gain running for energy reconstruction

Don Copage, Kansas
16
geometry dep. energy correction

• correction as function of ? and ? determined from
  p10 single electron MC (see D0-note 4023 Sabine
  Crepe-Renaudin)
• verification of geometry dependent energy
  correction on p11 data
• No electron between 1.2lt?detectorlt1.4.
• new correction factors for IC-region
• important correction
• reduced after fixes in ICR sampling weights?

?Emean vs. ?detector
Anne-Marie Magnan, Grenoble
17
geoEcor effect on Z-mass
no corrections
ECCC corrections only
ECCCICR corrections
SIGNAL128
SIGNAL129
SIGNAL122
NEW m0 90.25 0.5 GeV sigma 4.81 0.5 GeV
OLD m0 89.39 0.5 GeV sigma 5.15 0.6 GeV
WITHOUT m0 87.07 0.5 GeV sigma 4.98 0.5
GeV
Anne-Marie Magnan, Grenoble
18
Z?ee calibration
Alexis Cothenet, Marseille

• comparison of p10 and p11data
• same MC sample p10.11
• geometric corrections applied

p11 with increased trigger coverage to the EC
19
Z?ee calibration
Alexis Cothenet, Marseille
20
Calibration Coefficients
Alexis Cothenet, Marseille

• to be investigated check calorimeter operation,
  accumulate more statistics, split in run ranges

21
J/Psi with the road method
Florian Beaudette, Jean-Francois Grivaz, Orsay

• identify soft electrons searching for
  em-calorimeter depositions behind a track (see
  D0-note 4032)
• search of J/Psi events in p10 data
• tight cuts emfr gt 0.85 0.6ltE/plt1.05 E1,2,3gt0
  pTgt12.5 GeV pTassymlt0.6

J/Psi signal visible with opposite sign
tracks mass reconstruction from tracks
22
Efficiency for J/Psi selection
Florian Beaudette, Jean-Francois Grivaz, Orsay

• efficiency for 1 tight and 1 loose electron
  80 ? 0.97 (from MC)77
• mis-id probability (from K0? ?? decay) 2 ? 1
  in data 3 ? 1 in MC
• road method on L3 to be activated
• study with p11
• certification of road in EC

Jan Stark, Grenoble
23
low E resonance with scone

• resonance reconstruction from e with pT gt 2 GeV
  (hmtrx lt 60, iso lt 0.5)
• whole mass spectrum visible (combined p10, p11
  data)

Oleg Kouznetsov, Grenoble
24
? reconstruction
Oleg Kouznetsov, Grenoble

• extraction of ? peak (30 events)
• ? calibration at low energy
• but reduced low energy resonance signal since
  summer, due to absence of triggers!
• low ET trigger completely prescaled due to noise
• 2EM_2MD5 trigger went to 2EM_2MD7
• improvement from CEM(2,5)L3EM(2,7)
• (Ulla Blumenschein)
• reactivate CEM(2,2.5)

25
Outlook

• full p11 certification only sensible with
  improved MC!
• to come
• effect of various 0-suppression thresholds on
  em-reconstruction
• study of em-identification algorithms using
  CellNN objects
• calorimeter resolution determination
• not shown preparation and verification of
  tmb-nail and tmb-tree were done (Junjie Zhu,
  Maryland and Eric Thomas, Marseille)
• p13 improvements from tracking, corrected
  geometries and alignments