light jet energy scale from Wjj
See
•D. pallin: Tatra workshop ATLAS phys workshop (Rome) Top meetings
http://agenda.cern.ch/askArchive.php?base=agenda&categ=a041267&id=a041267
s4t4/transparencies
http://agenda.cern.ch/askArchive.php?base=agenda&categ=a044738&id=a044738
s11t4/transparencies
http://agenda.cern.ch/askArchive.php?base=agenda&categ=a057774&id=a057774
s1t3/moreinfo
• J schwindling Top meetings
http://agenda.cern.ch/askArchive.php?base=agenda&categ=a057774&id=a057774
s1t6/transparencies
http://agenda.cern.ch/askArchive.php?base=agenda&categ=a061436&id=a061436
s0t1/transparencies
http://agenda.cern.ch/askArchive.php?base=agenda&categ=a062180&id=a062180
s0t1/transparencies
1
Wjj sample in lepton+jets events

Select an enriched W sample from ttbar (lepton+jets) events

Selection with 1 or 2 btag

typically 3000 W/fb-1 (2btag, _btag=0.6)
x2 if 1 btag
( 80%-90%) purity (1 or 2 btag)
n
j1
W1
j2
b1
W2
t1
t2
l(e,m)
b2
Different Selection if 1btag


W mass shift linked to JES
Jets energies between 40 and 400 GeV
Rome AOD 500pb-1
2
Method 1 (D.Pallin)

Principle to obtain


E 2  (E 2 )  E 2




)
Constraint 1 Mjj=MWPDG
Constraint 2 the calibration function is universal
E1   ( E1 )  E1

E parton
  E jet
jet
E
M
PDG
W
 1 2 MW
part
with
E
 i  i jet
Ei
No hypothesis on function , no MC inputs
build W mass distributions in bin of jet energies
extract the peak values for each bins (n bins)
Deduce (E) from MW(E)
3
Method 1 W mass distributions
MW
MW
Rome AOD ‘cone 04’
MW
Ejet (GeV)
4
Method 1: Deducing (E) from MW(E)

Extraction of n values (Ebin)
Several possible methods




M PDG
  bini binj M W
W
chi2 fit
Iterative method
(equalisation of MC/DATA distribution)
Results (R) reproduce well the initial
calibration (theo) as function of E
EPart / E

1% precision with 1fb-1
E
5
Method 1: BUT

the calibration function as function of E is not the true calibration
EPart / E
E_jet / E_part % E with MC calib=1



Eparton

Bias is within 1% above 40 GeV
But need to be corrected
huge effect below 40 GeV
E recons
How to correct?
origine of bias : purely statistical from :
Ejet shape+ Ejet resolution
negligeable if resolution(E)<<E
important when resolution(E)~E
easy to compute if shape and resolution known


Shape given since W sample is 80% pure
Resolution could be computed from W width (or obtain from Z,+jets
samples)
6
Method :1 ongoing work

extraction of jet energy resolution from W width

evaluations of systematics
transportation of calibration to other samples
JES from 20 GeV
Increase size of initial sample

Note in preparation



7

Idea: generate mjj template
histograms by smearing W  qq
quarks with:
 Expected angular resolution
 b x expected energy resolution
 energy correlation between jets
 energy scale 
h: 297 / √E  11 mrad
j: 224 / √E  10 mrad
Expected angular resolution
Equark (GeV)
Expected Energy resolution
Fitted on Rome ttbar sample 4100

Idea more or less « à la CDF »
Will be presented in jet/ETmiss
meeting in Trigger/Physics week
end of May
s(Ej – Eq) (GeV)

Resolution (mrad)
Method 2: templates (J Schwindling)
s(E) = 3.8 GeV + 0.063xE
Eq (GeV)
Jet - quark
s(E1)
10.7
s(E2)
9.1
s(E1)  s(E2)
14.0
s(E1+E2)
16.5
(Error bars
~ 0.2 GeV)
+ 18 %
8
Meth2 :Expected performance


Fit both energy scale  and resolution b
With enough statisitics, can do this as a
function of energy (and/or h ?)
Injected 
1%
Reconstructed 

Example: apply templates on independent set
of smeared quarks with:
 Different energy scale(s)
 Different energy resolution
Injected resolution
3.0 + 0.080xE

Results (templates  smeared quarks):

Can fit average energy scale to ~0.4% with
1fb-1
 With more statistics (10 fb-1), could fit
energy scale as a function of E
 In principle allows also to check energy
resolution as a function of E
Fitted resolution
Default resolution
3.8 + 0.063xE
9
Meth 2: Towards real data
PT cut = 40 GeV


Try to fit the templates (smeared
PYTHIA quarks) on jj pairs from
MC@NLO + Geant 4 + reconstruction
(Rome sample 4100)
All jj combinations
Only 2 light jets
Only 2 light jets +
150 < mjjb < 200
Event selection (tt  jjb lnb):
 1 isolated lepton with PT > 20 GeV
 ETmiss > 20 GeV
 2 b-tagged jets with PT > 40 GeV
 2 and only 2 light jets with PT > 40
GeV
 150 GeV < mjjb < 200 GeV
 purity ~ 83 %, ~ 1200 jj pairs for 500
pb-1

Expect
 energy scale (Ejet / Eq) ~ 0.96 – 0.97

b=1
10
Meth 2: Conclusion / questions




Results (templates  sample 4100)

 = 0.958 ± 0.005

b = 1.07 ± 0.05
Fit of  seems very stable: ~ independent of qq smearing,
~ independent of mtop (± 0.4 % uncertainty on Escale for Dmtop
= ± 5 GeV), independent of combinatorial background
b sensitive to angular resolution, correlation between jet energies
… cannot claim to measure energy resolution
We need to understand better:
 Origin of energy correlation between jets
 Angular correlation ?
 Influence of underlying event
 E scale versus pileup
11