Philippe Doublet Roman Pöschl François Richard Thibault Frisson amp Jérémy Rouene LCWS Granada Sept 2011 Philippe Doublet LAL Plan Motivation Measurement method Efficiencies ID: 526759
Download Presentation The PPT/PDF document "Measuring the asymmetries of the top qua..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Measuring the asymmetries of the top quark at the ILC
Philippe DoubletRoman Pöschl, François Richard+ Thibault Frisson & Jérémy Rouene
LCWS Granada, Sept. 2011
Philippe Doublet - LALSlide2
Plan
MotivationMeasurement method
Efficiencies
Results
LCWS Granada, Sept. 2011
Philippe Doublet - LALSlide3
1. motivation
The top quark and flavor hierarchyGeography in Randall-Sundrum modelsTop to Z couplingsLCWS Granada, Sept. 2011Philippe Doublet - LALSlide4
The top quark and flavor hierarchy
Flavor hierarchy ? Role of 3rd generation ?
Top quark :
no
hadronisation
clean and detailed observations
Redo measurements of A
LR
and A
FB
with the top
LCWS Granada, Sept. 2011
Philippe Doublet - LALSlide5
Geography in Randall-Sundrum models
Higgs on IR brane for gauge hierarchy problemSM fermions have different locations along the 5th dimensionOverlaps leptons – Higgs in the 5th dimension generate good Yukawa couplings with O(1) localisation
parameters
t
0
R
e
0
A
µ
0
H
y (5th
dim
.)
0
π
R
UV
brane
(
M
Pl
)
IR
brane
(
TeV
)
bulk
LCWS Granada, Sept. 2011
Philippe Doublet - LAL
t
b
L
0Slide6
Top to Z couplings
Several RS models predict modified left
gZ(t
L) and right gZ
(tR) top
couplings
to Z (Z-Z
KK
mixing
, …)
Δ
g(
t
R
)/g(
t
R
)
Δ
g(
t
L
)/g(
t
L
)
SM 0,0
Djouadi
[1]
-34%,-1%
Hosotani
[2]
+18%,-7%
Gherghetta [3] -20%,-20%
Carena
[4]
0,-20%
LCWS Granada, Sept. 2011
Philippe Doublet - LALSlide7
2. Measurement
methodObservablesTop quark cross sectionMeasurement with the ILD detectorReconstruction within the ILD frameworkRequirements
LCWS Granada, Sept. 2011
Philippe Doublet - LALSlide8
Observables
σ(tt), ALR and AFB :
Semileptonic
decay mode :
tt
(
b
W
)(
b
W
)(
b
qq
)(
b
l
v
)
From
A
LR
and A
FB
, one
deduces
g
Z
(t
L) and gZ
(tR
) couplings
Allows
reconstruction of the top quark
Gives
top charge
LCWS Granada, Sept. 2011Philippe Doublet - LAL
A
LR =
(e- polar flip)
A
FB =
(top direction)
l = e, µSlide9
Top quark cross section
σ(tt) ≈ 600 fb at 500
GeV with 500 fb-1
Ntotal ~ 570k events
Semileptonic ~ 34%
Almost
background free ?
Major background =
other
top
channels
find
1
isolated
lepton
WW
no b quark
bb simple
topology
Major background : ZWW (Z
bb
)
≈ 8 fb
,
same
topologySmall but needs to be
subtractedLCWS Granada, Sept. 2011Philippe Doublet - LAL
500
Process
tt
bb
WWZZZWW
ALR (%)
36.762.998.8
31.089Slide10
Measurement
with the ILD detectorILD
optimised for
Particle Flow technique
(i.e.
reconstruct
every
particle
in a jet)
3.5 T B-
field
Performances :
Vertexing
:
σ
IP
= 5 µm (+) 10 µm/p(
GeV
)sin
3/2
θ
Tracking
:
σ
(1/
p
T
) < 5.10-5
GeV-1
Granular
calorimetry :
σE/E ~ 30%/√E
LCWS Granada, Sept. 2011
Philippe Doublet - LALSlide11
Reconstruction
whithin the ILD frameworkFull simulation is done
with the ILD detector
under GEANT4 (Mokka software)
« Objects
»
reconstructed
with
Particle
Flow
algorithm
(
Pandora
)
Data
used
:
samples
prepared
for the
LOIs
LCWS Granada, Sept. 2011
Philippe Doublet - LALSlide12
Requirements
ttbbqq
lv (l=e,µ)
Need
at
least 1 b jet (vertex)
Find
1 lepton (
tracking
)
Method
:
Find
a lepton
Force 4 jets
clustering
Find
at
least 1 (or 2) b jets
Form
the top
with
one b jet + 2 non-b jets
left
,
lepton charge
gives
the opposite
sign
of the topLCWS Granada, Sept. 2011
Philippe Doublet - LAL
e
-
e
+
neutrino
isolated
lepton
gives
top charge
b
b
2 light quarks = W
Hadronic
top to
reconstruct
θ
anti topSlide13
3. efficiencies
Identification of leptonsIsolationEfficiencies and purities of the selected leptonEfficiencies :
angular and energeticB tagging
LCWS Granada, Sept. 2011
Philippe Doublet - LALSlide14
Isolation
In reconstructed events, look
at the true
(MC) lepton :Events
forced to 4 jetstt
bbqqlv
: 4 jets + 1 lepton
Define
:
z =
E
lepton
/
E
jet
x
T
=
p
T
/
M
jet
Lepton
is
:
Leading
(
high
z)At high
pTNot
isolated
optimise cuts on z and
xT
True
lepton
embedded
inside a jetLCWS Granada, Sept. 2011
Philippe Doublet - LALSlide15
Results
of isolation cuts
Leading
Not isolated (lepton from b)
Large
p
T
Kinematic
limit
of
p
T
=
M
jet
/2
Blue = leptons in full
hadronic
top events = leptons from b
Cut
here
0.6
0.25
Red = leptons in
semileptonic
top events
Philippe Doublet - LAL
LCWS Granada, Sept. 2011Slide16
Efficiencies
: angular and energeticLCWS Granada, Sept. 2011
Philippe Doublet - LAL
Effiencies under control : Tracking worse in very forward regions
Leptons with small energies are suppressed by isolation cuts
Good efficiency with
~ full angular coverage
Efficiency = 87.9%
Contamination = 0.3%Slide17
B tagging
Vertex detector measure offset, multiplicity and mass
of jets to separate b from c decays
LCWS Granada, Sept. 2011
Philippe Doublet - LAL
Limitation of B tagging = B decay length
Good angular coverage
4 jets
2 highest
Btag
=
b
1
& b
2
2 “light” jets = W
B
Interaction pointSlide18
4. Results
Top reconstructionCross section and ALRProblem with the top reconstructionOrigin of the problemPrecisions reachedConclusions and prospects
LCWS Granada, Sept. 2011Philippe Doublet - LALSlide19
Top reconstruction
2 top candidates : (b1 + W) or (b2 + W)Retain candidate with minimald² = (
Mcand-M
t)²/σ
mt² + (
E
cand
-
E
beam
)²/
σ
Et
² +
(
M
W
rec
-M
W)²/
σ
mw
²
LCWS Granada, Sept. 2011
Philippe Doublet - LAL
σ
W
= 4.2
GeV
σ
t
=
7.1 GeV
~ 25 MeV
precision on MtSlide20
Cross-section and A
LRσ = N/(εL), L = 500fb-1After background suppression :Efficiency
= 72.7 % + Contamination = 4.6 % (
mostly full
hadronic top pairs)
σ(tt
SL
)
unpol
.
= 159.4 fb
Whizard
: σ(tt
SL
)
unpol
.
= 159.6 fb (-0.1%)
P(e-e
+)= (±80%, 0)
Δσ
/
σ
= 0.39% (stat.)
A
LR
= 0.435
ALR = 0.37 expected…
Whizard problem ?However, interest lies in relative
uncertaintyP(e-
e+)= (±80%, 0)
ΔALR
/ALR = 1.24% (stat.)
LCWS Granada, Sept. 2011
Philippe Doublet - LALSlide21
Problem
with the top reconstructionLCWS Granada, Sept. 2011Philippe Doublet - LAL
Migration
No migration ?
Relative errors : -5.2% (
A
FB
t
R
)
-40.4 % (
A
FB
t
L
)
1.1 % (
stat.)
Angular distributions for e
-
L
and e
-
R
eventsSlide22
Solving
the problemIs it due to the reconstruction ?Cut on the quality of the candidate (particle flow)
Efficiency in e-
L : x60%
relative systematics
:
40%
20 %
Is
is
intrinsic ?
Effect of
helicity
structure of the decays
Ambiguous solutions
Seen with
partonic
reco
.
LCWS Granada, Sept. 2011
Philippe Doublet - LAL
d² = (
M
cand
-M
t
)²/
σ
mt
² + (
E
cand-Ebeam
)²/σEt² +
(MWrec-M
W)²/σ
mw²
quality of the candidate
“
parton limit”Slide23
Precisions reached
LCWS Granada, Sept. 2011Philippe Doublet - LAL
Pe
- / Pe+
(80% / 0)A
LR
A
FB
t
R
A
FB
t
L
Q
Z
tL
Q
Z
tR
stat. error
1.3%
1.2 %
1.4 %
1.0 %
1.9 %
Possible to probe some RS models with M
KK
~ 2.8
TeV
up to 25
TeV
Correction on
AFBtL
= dominant systematic (reco. + intrinsic) Good PFA + b tagging are essential
20% correction on AFBtL can be done on a well tuned MCSlide24
Conclusion and prospects
Impact of detector & reconstruction performances on a complex channel : lepton + 4 jets with 2 b jetsFinal efficiency = 72.7%Contamination = 4.6% (Major backgrounds are other top channels)
σ and A
LR can be known at 0.4% and 1.3% statistical uncertainty (
systematics
guaranteed small due to large purity)
Problem in reconstructing the direction of the top
Reconstruction needs improvements or leads to efficiency losses
Intrinsic problem with
A
FB
t
L
needs excellent Monte Carlo
A
FB
t
R
/L
known with 1.2/1.4% statistical uncertainty
Study of A
FB
to enter the DBD for the ILD in 2012
LCWS Granada, Sept. 2011
Philippe Doublet - LALSlide25
5. Additional
materialTop physics : LHC and ILCTop couplings : bibliography
LCWS Granada, Sept. 2011
Philippe Doublet - LALSlide26
Top
physics : LHC and ILCLC 1 pb, LHC 1nb but for gluon couplings onlyVery good s/b at ILC and energy/momentum conservation allows to reconstruct modes with a neutrinoMt and
Gt with ≈50
MeV error, 0.4% on cross section LC unique to measure t
R and tL
Z couplings at % (ND>4) LHC > 10 times worse
ILD
LOI
LCWS Granada, Sept. 2011
Philippe Doublet - LALSlide27
Top
couplings : bibliography[1] : Djouadi et al., Nuclear Physics B, Volume 773, Issues 1-2, 25
June 2007, Pages 43-64 [2] : Hosotani
et al., Prog. Theor
. Phys. 123 (2010), 757-790[3] : Cui, Gherghetta
et al.,
arXiv
:1006.3322v1 [hep-ph]
[4] :
Carena
et al.,
Nuclear Physics B
Volume 759, Issues 1-2, 18 December 2006, Pages 202-227
LCWS Granada, Sept. 2011
Philippe Doublet - LAL