NP workshop

1
From p11 to p13
Missing ET distributions Warm Zones, CH, Fake
Jets, Coherent
treatment of Calo for jet/met/e.m. objects for
p13 and beyond Run selections
2
MET in root-tuple (p11)

• METx, METy and MET distributions are not stable
• Run selection smoothen distributions but only
  60 efficiency (with METC)

3
EM and FH channels

• Cut 3s in EM 4s in FH

4
Channel noise

• 1st line raw noise/ADC 2nd cal weights
  3rd noise/GeV
• lt 40 MeV gt100 MeV

R. Zitoun
5
Better to cut on x-y or RMS ?

• root-tuple MET (only cells above 2.1s and Egt100
  MeV ? METC)
• only cells above 3/4/6 s in EM/FH/CH
• no CH in MET
• 3/4/6 s cut and no CH
• Can we understand these shifts / offsets?

6
Warm Zone Identified
Warm zone present in the data since a long time
(before summer 2002) Visible in all e.m.
layers. Not really visible in had. Layer. Low
energy cells. Effect checked on MET by removing
(opposite in phi)
7
Warm Zone Identified
Characteritic of a BLS crate problem fix done.
To be checked off-line
8
Removing Warm Zone from METx,y
Effect on mean, one entry per run. Before A
fter
9
Removing Coarse Hadronic in METx,y
Effect on mean one entry per run. With
CH Without
10
Removing Coarse Hadronic in METx,y
Effect on rms one entry per run. With
CH Without
11
Removing Coarse Hadronic in METy
mean METy, one entry per run. With CH Without CH
rms METy, one entry per run. With CH Without CH
12
Effect of Coarse Hadronic Layers and negative
cells
P13 study Emmanuel Busato
Current status of the algorithms
Jets are constructed with a list of positive
energy towers. These towers are made of
positive and negative energy cells.
13
Cone R0.5
Current algo
Without CH layers
Without CH layers and neg cells
Cone R0.7
Without CH layers and neg cells
Without CH layers
Current algo
14
Coherent Treatment for p13
Preliminary version available at www-d0.fnal.gov/
gregorio/energy.ps
15
Proposal for p13
All analyses based on p13 should NOT use the
unclustered energy in the CH (layers 15,16,17)
?They should use the CH cells only when they
are in towers which are part of a jet satisfying
the Jet-Met Criteria. To a good approximation
for the unclustered energy, use only positive
energy cells in layers lt 15.
16
Technically realized by
1) Computing an initial missing transverse energy
MET-0 (x,y,z components) using all cells of the
CAL block in layers between 1 and 14 and of
positive energy or negative energies, provided
that they belong to a tower of global positive
energy. Here a tower is constructed as usual with
all layers. 2) Adding to MET-0 the 4-momentum of
each jet satisfying the Jet-Met quality criteria
weighted by its coarse hadronic fraction. In this
way we add only the CH energy of the "good" jets,
since the rest of the energy has already been
included in the inclusive computation and since
we don't want to include the CH energy of "fake"
jets. (there is the small bias due to the
difference between pTch and pTjet direction,
negligible compared to bias due to full CH
effect, and which can be corrected if needed) 3)
Correcting the missing transverse energy
according to the standards jet and e.m.-objects
corrections, and for the presence of good
muons.
17
100 MeV or 0 Mev Threshold ?
one entry per run. 0 MeV 100 MeV
Mean MET-x (no CH)
100 MeV protects against some patho- logies.
Useful at the beginning, thr.will be set to 0 MeV
in p14 ? MET-0 (no CH)
0 MeV 100 MeV
RMS MET-x (no CH)
18
MET-0 (noCH) vs METC (withCH)
one entry/run. METC-y MET-0-y
Mean MET-y
Mean MET-x (no CH)
MET-0 is much more stable than METC. Main
effect is removal of CH.
METC-y MET-0-y
RMS MET-x (no CH)
RMS MET-y
19
Good Run Definition
Only Runs with gt 1000 events are considered.
(187 runs, p11) Define the average shift
MET-xy sqrt (ltMET-xgt)2 (ltMET-ygt)2 For a
GOOD RUN 1) Require MET-xy lt 3 GeV in at least
75 of the root-files of the run. In addition
all the root-files must have MET-xy lt 5 GeV 2)
Require mean scalar ET between 90 and 160
GeV. With METC We keep 123 good
runs ( 60 of the total statistics, about 8
pb-1) With MET-0 We keep 159
good runs ( 85 of the total statistics, about
11.2 pb-1) Accepting all runs with MET-0-xy lt 5
GeV (loose condition) We keep 171
good runs ( 94 of the total statistics, about
12.5 pb-1)
20
August-October Run Selection
one entry per run. METC-x MET-0-y
Mean MET-x (no CH)
159 run selected
METC-x MET-0-y
RMS MET-x (no CH)
21
Conclusions
With the new proposed treatment (same cells used
up to layer 14 for MET, Jets, e.m. objects.
Rejection of unclustered energy in CH)
?improve Missing ET resolution ?increase the
number of good runs ? higher statistics ?allow to
find smaller calorimetric effects (shift still
there) ?diminish importance of fake jets, since
only the non-CH part of their energy is used in
the missing ET calculation. ?allows to
concentrate further p13 efforts on jets ?set the
standard for future improvements for p14
enhanced NADA in CH layers
new T42 algorithm to reduce
low energy noise