tevatron cdf top quark massrencontres de moriond ew, la thuile, march 19, 2014 ... top quark mass...
TRANSCRIPT
1✬
✫
✩
✪
Rencontres de Moriond EW, La Thuile, March 19, 2014
CDFTEVATRON
Top Quark Mass
L. BrigliadoriUniversity of Bologna
On behalf of the CDF and DØ Collaborations
OUTLINE• The Top Quark
• Measurement Strategies.
• Recent Results.
• Mtop World Average
• Summary and Conclusions.
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
2✬
✫
✩
✪
CDF The Top Quark
• Observed in 1995 at Fermilab....... not a big surprise (...if we trust the SM...)
• ... but Mtop striking large!
• Why (keep on) to measureMtop ?
– Inside the SM:
∗ Test SM predictions
∗ Test SM consistency
– Beyond the SM:∗ Constraints on New Physics
∗ Hints on EWSB?
(GeV)topm
(G
eV)
Wm
< 127 GeV
H
115 < m
< 1000 GeV
H
600 < m
155 160 165 170 175 180 185 190 195
80.3
80.35
80.4
80.45
80.5
top (2009), m
W68% CL (by area) m
top (2012), m
W68% CL (by area) m
& direct Higgs exclusion)top
, mW
LEPEWWG (2011) 68% CL (excluding m
March 2012
[GeV]tm
160 165 170 175 180 185 1902 χ
∆0
1
2
3
4
5
6
7
8
9
10
σ1
σ2
σ3 measurementt
SM fit w/o m
measurementsH
and Mt
SM fit w/o m
ATLAS measurement [arXiv:1203:5755]tkinm
CMS measurement [arXiv:1209:2319]tkinm
Tevatron average [arXiv:1305.3929]tkinm
[arXiv:1207.0980]tt
σ obtained from Tevatron tpole
m
G fitter SM
Sep 13
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
3✬
✫
✩
✪
CDF Top Quark Signatures
At the Tevatron mainly tt̄ production via strong interaction
∼ 85%
+ +
∼ 15%
In the SM : BR(t → Wb) ≃ 100% ⇒ tt̄ final states defined byW s’ decays
2 high pT charged leptons,2 b-jets, /ET
4 q-jets,2 b-jets
1 high pT charged lepton,2 q-jets, 2b-jets, /ET
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
4✬
✫
✩
✪
CDF Top Mass Challenges
• σtt̄ / σinel ≃ 10−10 !!!
• ... Event Selection :
– Triggers– b-tagging algorithms.– High ET and central (|η| ≤ 2) Jets.– Lepton Id (Dilepton, Lepton + jets).
• Reconstruction :
– Measure “Jets” and not partons
Need corrections to obtain parton energy
⇒ JetEnergyScale. σJES/JES ≈ 2% to 6%
⇒ Important contribution toσMtop(syst)
– Jets-to-partons assignments
Which jet comes from which particle?Combinatoric problem!
– Undetectedν ’s (Dilepton, Lepton + jets).
Need assumptions. Multiple solutions.
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
5✬
✫
✩
✪
CDF Measurement Strategies
• Matrix Element (ME)
– Define the probability, Pev , that the observed kinematics,~y, arise from possible signal or bkgkinematics~x at parton level :
∗ dσ(~x) LO differential x-section of a final state~x at parton level.Depending onMtop for tt̄ events, but not for bkg.
∗ W(~y, ~x) “Transfer function”, i.e. probability to measure the observedset of variables~y, given~x at parton level. Depends onJES.
∗ ftt̄ Fraction of signal events expected in the data.
– Maximize Lsample∝∏
events
Pev(~y, ftt̄, Mtop) evaluated for observed data
• Template Method
– Consider a set of observables,~x, sensitive toMtop.Evaluate and plot the set for each event⇒ “Templates”
– Maximize a likelihood whereobserved distributionsare compared to expectations for differentMtop andsignal fractions,ftt̄.
Lsample ∝∏
events
∏
~x
Lshape(xi|ftt̄, Mtop)
]2 [GeV/crectm
100 150 200 250 300
)2F
ract
ion
of E
vent
s/(2
.5 G
eV/c
0
0.02
0.04
0.06
0.08
0.1
2 = 167.5 GeV/ctop M
2 = 172.5 GeV/ctop M
2 = 177.5 GeV/ctop M
JES)∆,top
| Mrect
(msP
templatesrect mtt
JES = 0.0)∆ 2 tags events (≥
)-1 - Preliminary (9.3 fbtopCDF Run II - All Hadronic M
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
6✬
✫
✩
✪
CDF Measurement Channels
Lepton + Jets
Reasonable Bkg,Good Statistics....
The Golden Channel!
• Jets-to-Partons assignment ambiguity
• Well reconstructed kinematics (butpνz
ambiguity)
Dilepton
The cleanest sample...The smallest statistics
• Small combinatoric problem
• Underconstrained kinematics (2 undetectedν’s)
All-Jets
Huge QCD Bkg...Large Statistics
Challenging!
• Need “fine tuned” selections to obtain goodS/B
• Large Jets-to-Partons assignment ambiguity
• Fully reconstructed kinematics
MissingET + Jets
• Selection defined to be complementaryto other channels.
• Mostly L + Jets with undetected lepton
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
7✬
✫
✩
✪
CDF Dilepton
• DØ , Matrix Element, 5.4 fb−1
• Phys. Rev. Lett. 107, 082004 (2011).
• ee, eµ, µµ channels
Mtop = 174.0 ± 1.8 (stat) ± 2.4 (syst)GeV
(GeV)tm160 170 180 190
max
tot
/Lto
tL
0
0.5
1
-1D0 5.4fb
• DØ , Template by NWA, 4.3 + 1.0 fb−1
• Phys. Rev D86, 051103(R) (2012)
• Neutrino Weighting Algorithm : templates based on an event weightfor possible solutions of underconstrained kinematics.
• ee, eµ, µµ channels, selection similar to ME analysis.
• JES calibration from Lepton + Jets analysis is used.
• Combined with NWA + Matrix Weighting @ 1 fb−1
Mtop = 174.0 ± 2.4 (stat) ± 1.4 (syst)GeV
(GeV)
wµ120140160180200 (GeV)
tm
160170
180190
200
Pro
babi
lity
dens
ity
00.10.20.30.40.50.6
-310×
WTνDØ,
=30 GeVwσ
d)
• Combination :Mtop = 173.9±1.9 (stat)±1.4 (syst)GeV
σMtop/Mtop ≃ 1.4%
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
8✬
✫
✩
✪
CDF Dilepton
• CDF , Templates, 9.1 fb−1
• NewPreliminary result with full CDF dataset(January 2014, CDF conference note 11072)
• “Hybrid” variable method to reduce JES uncertainty :
– Templates byMefft = w · Mreco
t + (1 − w) · Maltt
– Mrecot sensitive to trueMtop. Defined by NWA.
– Maltt less sensitive toMtop, but not based on jet energies.
– w = 0.7 :defined to minimize expected (stat + JES) uncertainty
• Two independent samples :0-tag, ≥ 1-tag
Mtop = 170.80 ± 1.83 (stat) ± 2.69 (syst)GeV
σMtop/Mtop ≃ 1.9%
0-tag fitted templates
≥ 1-tag fitted templates
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
9✬
✫
✩
✪
CDF Lepton + Jets
• DØ, Matrix Element, 2.6 + 1.0 fb−1
• Phys. Rev. D84 (2011) 032004.
• Events with 1 highpT e /µ, ≡ 4 Jets, large/ET , ≥ 1 b-tag
• For each event evaluate ME-based probability as aweighted sum over possible parton-jet combinations
• Dependence onJES given by Transfer FunctionsW(~y, ~x)
• SimultaneousJES calibration by jets assigned toW boson.
(GeV)xE0 50 100 150
)-1
=1)
(GeV
JES
,k y,E x
(Eje
tW 0
0.02
0.04
0.06
(GeV)yE 30 60 90 120
| < 1η0.5 < |(b) DØ
JetE Transfer Functions
GeVtm170 172 174 176 178 180 182
JES
k
0.98
0.99
1
1.01
1.02
1.03
1.04
1.05
1.06-1DØ, 2.6 fb
1sd
2sd
3sd
Mtop = 176.01 ± 1.01 (stat) ± 1.29 (syst)GeV
Combined with 1 fb−1⇒ Best DØ measurement
Mtop = 174.94 ± 0.83 (stat) ± 1.24 (syst)GeV
σMtop/Mtop ≃ 0.9%
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
10✬
✫
✩
✪
CDF Lepton + Jets
• CDF, Templates, 8.7 fb−1
• Phys. Rev. Lett. 109 (2012) 152003.
• 3D templates :mrecot vs m
reco(2)t vs mjj
• mjj used for in situ JES calibration
Mtop = 172.85 ± 0.71 (stat) ± 0.85 (syst)GeV
σMtop/Mtop ≃ 0.6%
Best single measurement from Tevatron
)2 (GeV/crecotm
120 140 160 180 200 220 240
)2 (
GeV
/cjj
m
50
60
70
80
90
100
110
0.0002
0.0004
0.0006
0.0008
0.001
0.0012
)2 = 172.0 GeV/ctop
lepton + jets, 1-tag Signal probability (M ) 2 = 172.0 GeV/ctop
lepton + jets, 1-tag Signal probability (M
mjj vsmrecot
)2
(GeV/cjjm50 60 70 80 90 100 110 120
Arb
uni
ts
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
2-tag)cσ (JES∆-1.0
)cσ (JES∆0.0
)cσ (JES∆+1.0
2-tag
mjj dependence onJES
)2
(GeV/crecotm
100 150 200 250 300 350
)2E
vent
s/(5
GeV
/c
0
50
100
150
200
250
300
350
400CDF II Preliminary
)-1Data (8.7 fbSignal+BkgdBkgd onlyAll
)2 (GeV/cm50 60 70 80 90 100 110 120
)2E
vent
s/(1
.5 G
eV/c
0
50
100
150
200
250
300
350CDF II Preliminary
)-1Data (8.7 fbSignal+BkgdBkgd onlyAll
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
11✬
✫
✩
✪
CDF All-Hadronic
• CDF, Templates, 9.3 fb−1
• NewPreliminary result with full CDF dataset(February 2014, CDF conference note 11084)
• Tuned event selection based on Neural Net,b-tag
• Data driven bkg modeling
• S/B about1/1 for ≥ 2-tag events.
• In each event reconstruct (χ2 minimization) :— a “top mass”, mrec
t
— a “W mass”, mrecW
⇒ JES calibration
]2 [GeV/crectm
100 150 200 250 300
)2F
ract
ion
of E
vent
s/(2
.5 G
eV/c
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
2 = 167.5 GeV/ctop M
2 = 172.5 GeV/ctop M
2 = 177.5 GeV/ctop M
JES)∆,top
| Mrect
(msP
templatesrect mtt
JES = 0.0)∆1 tag events (
)-1 - Preliminary (9.3 fbtopCDF Run II - All Hadronic M
]2 [GeV/crecWm
20 40 60 80 100 120 140 160 180 200 220
)2F
ract
ion
of E
vent
s/(2
.5 G
eV/c
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
JES = -2.0∆
JES = 0.0∆
JES = 2.0∆
JES)∆,top
| MrecW
(msP
templatesrecW mtt
)2 = 172.5 GeV/ctop
1 tag events (M
)-1 - Preliminary (9.3 fbtopCDF Run II - All Hadronic M
]2 [GeV/crectm
100 150 200 250 300
)2E
vent
s/(5
.0 G
eV/c
0
100
200
300
400
500 1-tag events≥
)-1 Data (9.3 fb
t Fitted Bkg + t
Fitted Bkg
)-1 - Preliminary (9.3 fbtopCDF Run II - All Hadronic M
Mtop = 175.07±1.19 (stat)±1.56 (syst)GeV
σMtop/Mtop ≃ 1.1%
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
12✬
✫
✩
✪
CDF MissingET + Jets
• CDF, Templates, 8.7 fb−1
• Phys. Rev D88 (2013) 011101
• Require large/ET , no tight lepton
⇒ independent of All-Had, L+jets, Dilepton
• Exploiting NN selection and bkg modelingsimilar to All-Had analysis
• /ET treated as due toW → ℓν with “lost” charged leptonin event reconstruction
• Same technique used in CDF L+jets :
– 3D templates usingmrecot , mreco(2)
t , mjj
– mjj used for in situ JES calibration
Mtop = 173.93 ± 1.26 (stat) ± 1.36 (syst)GeV
σMtop/Mtop ≃ 1.1%
NN output 0 0.5 1
Eve
nts
0
50
100
150
200
Data
)2 = 172.5 GeV/ct
Signal (M
Background
)-1CDF Run II Preliminary (8.7 fb
2tag MET+Jets Events
Neural Net Output
)2 (GeV/ctopM170 172 174 176 178
)2 (GeV/ctopM170 172 174 176 178
) cσ (JE
S∆
-0.6
-0.4
-0.2
0
0.2
0.4
0.6 log(L) = 0.5∆
log(L) = 2.0∆
log(L) = 4.5∆
-1CDF II Preliminary 8.7 fb
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
13✬
✫
✩
✪
CDF Mtop Systematic Uncertainties
⇐ example from CDF Dileptons
• Precision on the Top Mass measurementsnow limited by systematic uncertaintiesin all channels.
• JES uncertainty greatly reduced especially byin situ calibration techniques⇒ partially statistical
• CDF and DØ Collaborations performed a jointeffort in the past years in order :∗ to define a common way to evaluate
systematics
∗ to avoid possible “double counting” of someeffect
∗ to evaluate correlations among differentmeasurements
∗ to study possible neglectedsources of uncertainties
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
14✬
✫
✩
✪
CDF Tevatron Top Mass
CDF + DØ, March 2013 (arXiv:1305.3939)
Mtop = 173.20±0.51(stat)±0.71(syst)GeV
Individual channels
Dilepton Mtop = 170.0 ± 2.1GeV
Lepton + Jets Mtop = 173.2 ± 0.9GeV
All-Hadronic Mtop = 172.7 ± 1.9GeV
/ET + Jets Mtop = 173.8 ± 1.8GeV
• Best results of each experiment in each channelfrom Run I and Run II combined.
• Recent updates from CDFnot included yet
• All correlations taken into account.
• Good agreement among resultsfrom individual channels
• Mtop known at 0.50% (March ’13).
• Precision now limited by systematic uncertaintiesin all channels.
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
15✬
✫
✩
✪
CDF Mtop WORLD AVERAGE
The ATLAS, CDF, CMS and DØ Collaborations just approved thevery first Tevatron + LHC Mtop combination
Mtop = 173.34 ± 0.76GeV
[GeV]topm165 170 175 180 185
1
17
LHC Sept. 2013 0.88)± 0.26 ± (0.23 0.95±173.29
Tevatron March 2013 (Run I+II) 0.61)± 0.36 ± (0.51 0.87±173.20 prob.=93%2χ / ndf =4.3/102χ World comb. 2014 0.67)± 0.24 ± (0.27 0.76±173.34
-1 = 3.5 fbint L
CMS 2011, all jets 1.23)± (0.69 1.41±173.49
-1 = 4.9 fbint L
CMS 2011, di-lepton 1.46)± (0.43 1.52±172.50
-1 = 4.9 fbint L
CMS 2011, l+jets 0.97)± 0.33 ± (0.27 1.06±173.49
-1 = 4.7 fbint L
ATLAS 2011, di-lepton 1.50)± (0.64 1.63±173.09
-1 = 4.7 fbint L
ATLAS 2011, l+jets 1.35)± 0.72 ± (0.23 1.55±172.31
-1 = 5.3 fbint L
D0 RunII, di-lepton 1.38)± 0.55 ± (2.36 2.79±174.00
-1 = 3.6 fbint L
D0 RunII, l+jets 1.16)± 0.47 ± (0.83 1.50±174.94
-1 = 8.7 fbint L
+jetsmiss
TCDF RunII, E
0.86)± 1.05 ± (1.26 1.85±173.93 -1 = 5.8 fbint L
CDF RunII, all jets 1.04)± 0.95 ± (1.43 2.01±172.47
-1 = 5.6 fbint L
CDF RunII, di-lepton 3.13)± (1.95 3.69±170.28
-1 = 8.7 fbint L
CDF RunII, l+jets 0.86)± 0.49 ± (0.52 1.12±172.85
-1 - 8.7 fb-1 = 3.5 fbint
combination - March 2014, LtopTevatron+LHC mATLAS + CDF + CMS + D0 Preliminary
) syst. iJES stat.total (Pre
viou
sC
omb.
∗ arXiv 1403.4427 [hep-ex]
∗ Tevatron : Run II data(up to 8.7 fb1)
∗ LHC : 2011 data(up to 4.9 fb1)
∗ Best single measurementin each channelfrom each experiment
∗13% improvement w.r.t.most precise singleCollider combination
∗28% improvement w.r.t.most precise single input
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
16✬
✫
✩
✪
CDF Mtop WORLD AVERAGE
ρEXP ρLHC ρTEVρCOL
ρCDF ρD0 ρATL ρCMS ρATL−TEV ρCMS−TEV
Stat 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
iJES 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0
stdJES 1.0 1.0 1.0 1.0 0.0 0.0 0.0 0.0
flavourJES 1.0 1.0 1.0 1.0 0.0 0.0 0.0 0.0
bJES 1.0 1.0 1.0 1.0 0.5 1.0 1.0 0.5
MC 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
Rad 1.0 1.0 1.0 1.0 1.0 1.0 0.5 0.5
CR 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
PDF 1.0 1.0 1.0 1.0 1.0 1.0 0.5 0.5
DetMod 1.0 1.0 1.0 1.0 0.0 0.0 0.0 0.0
b-tag 1.0 1.0 1.0 1.0 0.0 0.0 0.0 0.0
LepPt 1.0 1.0 1.0 1.0 0.0 0.0 0.0 0.0
BGMC† 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
BGData 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Meth 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
MHI 1.0 1.0 1.0 1.0 1.0 0.0 0.0 0.0
Multiplicative factor f [%]
0 20 40 60 80 100
[MeV
]to
p m∆
-250
-200
-150
-100
-50
0
50
100
150
200
EXPρ ×f
TEVρ ×f
LHCρ ×f
COLρ ×f
ALLρ ×f
ATLAS + CDF + CMS + D0 Preliminary
Multiplicative factor f [%]
0 20 40 60 80 100)
[MeV
]to
p(mσ ∆
-350
-300
-250
-200
-150
-100
-50
0
50
100
EXPρ ×f
TEVρ ×f
LHCρ ×f
COLρ ×f
ALLρ ×f
ATLAS + CDF + CMS + D0 Preliminary
• First combination between the two Colliders
• Big effort performed in order to :
∗ classify uncertainties
∗ define correlations
• Various correlation scenarios have been checked
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
17✬
✫
✩
✪
CDF Mtop WORLD AVERAGE
• Combine channels / experiments / colliders simultaneoslybut separately(i.e. correlated but different mtop fit parameters)
[GeV]topm170 172 174 176 178 180
1
16
LHC Sept. 2013 0.88)± 0.26 ± (0.23 0.95±173.29
Tevatron March 2013 (Run I+II) 0.61)± 0.36 ± (0.51 0.87±173.20
World comb. 2014 0.67)± 0.24 ± (0.27 0.76±173.34
LHC 0.87)± 0.26 ± (0.23 0.94±173.26
Tevatron 0.69)± 0.36 ± (0.46 0.91±173.41
CMS 0.94)± 0.29 ± (0.30 1.02±173.54
ATLAS 1.34)± 0.41 ± (0.31 1.43±172.70
D0 1.13)± 0.49 ± (0.79 1.46±174.62
CDF 0.68)± 0.45 ± (0.53 0.98±172.96
+jetsmissTE 0.71)± 1.05 ± (1.28 1.80±174.03
all jets 0.89)± 0.25 ± (0.65 1.13±173.35
di-lepton 0.98)± 0.19 ± (0.44 1.09±172.73
l+jets 0.68)± 0.26 ± (0.27 0.78±173.23
-1 - 8.7 fb-1 = 3.5 fbint
corr. obs. - March 2014, LtopTevatron+LHC m
ATLAS + CDF + CMS + D0 Preliminary
) syst. iJES stat.total (
Cor
rela
ted
Obs
erva
bles
Pre
viou
sC
omb.
∗ Very good agreementamongall fitted values
∗ ATLAS-CONF-2014-008CDF note 11071CMS PAS TOP-13-014DØ note 6416
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
18✬
✫
✩
✪
CDF NEW CDF Top Mass
• New CDF update:
∗ The CDF collaboration just approved anew combination of measurements.
∗ All channels now updated with full dataset.CDF public note 11080
)2 (GeV/ctM150 160 170 180 190 200
0
11
CDF March’07 2.66± 12.40 2.20)±1.50 ±(
CDF combination * 0.93± 173.16 0.74)±0.57 ±(
syst)± stat ±(
CDF-II MET+jets 1.85± 173.93 1.26)±1.35 ±(
CDF-II track 9.46± 166.90 2.90)±9.00 ±(
CDF-II alljets * 1.96± 175.07 1.56)±1.19 ±(
CDF-I alljets 11.51± 186.00 5.70)±10.00 ±(
CDF-II lepton+jets 1.11± 172.85 0.98)±0.52 ±(
CDF-I lepton+jets 7.36± 176.10 5.30)±5.10 ±(
CDF-II dilepton * 3.25± 170.80 1.83)±2.69 ±(
CDF-I dilepton 11.41± 167.40 4.90)±10.30 ±(
Mass of the Top Quark
(* CDF preliminary)March 2014
/dof = 5.6/7 (58%)2χ
Mtop = 173.16 ± 0.57(stat) ± 0.74(syst)GeV
= 173.16 ± 0.93GeV
Analysis0
0.2
0.4
0.6
for
all m
easu
rem
ents
i w
eigh
t
∑ �/
i
wei
ght
CDF-II lepton+jets
CDF-II MET+jets
CDF-II allje
ts
CDF-I lepton+jets
CDF-II dile
pton
CDF-I dile
pton
CDF-I allje
ts
CDF-II tra
ck
March 2014CDF Preliminary
Red box: positive weightsGrey box: absolute value of
negative weights
• Last word on Mtop from CDF
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
19✬
✫
✩
✪
CDF Conclusions
• The Fermilab TEVATRON was shut downin September 2011∼ 10 fb−1 of data collected by experiments during Run II.
• The Collaborations are finalizing measurements of theTop Quark Mass using full datasets
• A selection of more recent ones has been presentedin this talk. Full details in
∗ www-cdf.fnal.gov/physics/new/top/publicmass.html
∗ www-d0.fnal.gov/Run2Physics/top/
• Excellent results from all decay channels.Uncertainty now everywhere dominated by systematics.
• Results from individual experiments haveprecisions by far beyond Run IIa goal.
• Tevatron competitive with LHC.
• It’s time for Mtop World Average
)−1Integrated Luminosity (pb
210 310 410
2(t
otal
) G
eV/c
t M∆
1
10
CDF Results
Run IIa LJ goal (TDR 1996)
(syst) fixed∆, L(stat) scales as 1/∆
L(total) scales as 1/∆2 = 1GeV/c
t/MtM∆
CDF Top Mass Uncertainty(all channels combined)
−11 fb −12 fb −15 fb −110 fb
Now : σMtop≃ 0.93GeV
LHC + TEVATRON, March 2014
Mtop = 173.34 ± 0.76GeV
σMtop/Mtop ≃ 0.44%
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
20✬
✫
✩
✪
Backup
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
21✬
✫
✩
✪
CDF The Tevatron Experiments
CDF
DØ
• pp̄ collisions at 1.96 TeV (Run II, 2001-2011).
• Peak lumi≈ 4 · 1032cm−2s−1
• About 12 fb−1 delivered to experiments.
Acquired 10 fb−1 / experiment
• Collaborations :– Currently 400 + 400 members– 60 (CDF) + 70 (DØ) Institutions
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
22✬
✫
✩
✪
CDF Matrix Element Method
• Matrix Element (ME)
– Defineper-event probability by Leading Order ME of signal ( tt̄) and Bkg events as a functionof Mtop, JES and the expected fraction of signalftt̄ :
Pev (~y, ftt̄, Mtop, JES) = ftt̄ ·Ptt̄ (~y, Mtop, JES)+(1 − ftt̄) ·Pbkg (~y, JES)
Ptt̄ ∝1
N
∫
f(z1) f(z2) dz1 dz2︸ ︷︷ ︸
p.d.f .
W(~y, ~x, JES)︸ ︷︷ ︸
Transfer function
dσtt̄(~x, Mtop)︸ ︷︷ ︸
differentialx−section
– Pbkg totally analogous, butdσbkg(~x).
– Pev gives the probability for the observed event kinematics,~y, to arise from a signal or a bkgevent.
– N : Normalization factor
– f(z) : Parton density functions
– W (~y, ~x, JES) : Connect observed jets to partons. Give the probability forthe measuredjet momenta~y given corresponding parton momenta~x. Depend on the Jet Energy Scale.
– dσ(~x) : Include ME calculation and phase space. Depend onMtop for tt̄ events.
• Maximize sample likelihood L (~y, ftt̄, Mtop, JES) =∏
events
Pev
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
23✬
✫
✩
✪
CDF Template Method
• CDF, Lepton + Jets: χ2 expression formrecot (free parametersmreco
t , pfitT,i
andUfitj
)
χ2 =
(mjj − MW
)2
Γ2W
+
(mlν − MW
)2
Γ2W
+
(mjjb − mreco
t
)2
Γ2t
+
(mlνb − mreco
t
)2
Γ2t
+∑
i=l,jets
(pfitT,i
− pmeasT,i
)2
σ2i
+∑
j=x,y
(U
fitj
− Umeasj
)2
σ2j
• CDF, All-Hadronic : χ2 expression formrect (free parametersmrec
t and pfitT,i
)
χ2 =
(m
(1)jj
− MW
)2
Γ2W
+
(m
(2)jj
− MW
)2
Γ2W
+
(m
(1)jjb
− mrect
)2
Γ2t
+
(m
(2)jjb
− mrect
)2
Γ2t
+∑
i=jets
(pfitT,i
− pmeasT,i
)2
σ2i
∗ mjj , mlν : Invariant masses of dijet and lepton-neutrino systems
∗ mjjb, mlνb : Invariant masses of three-particle systems includingb-jets
∗ pmeasT,i
, σi : Measured transverse momenta of lepton, jets and uncertainties
∗ Umeasj
, σj : components of unclustered energy and uncertainties.
∗ MW , ΓW , Γt : Mass ofW boson and widths ofW and top quark
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014
24✬
✫
✩
✪
CDF Mtop WORLD AVERAGE
• Uncertainties
Input measurements and uncertainties in GeV
CDF D0 ATLAS CMS World
Uncertainty l+jets di-l all jet EmissT
l+jets di-l l+jets di-l l+jets di-l all jet Combination
mtop 172.85 170.28 172.47 173.93 174.94 174.00 172.31 173.09 173.49 172.50 173.49 173.34
Stat 0.52 1.95 1.43 1.26 0.83 2.36 0.23 0.64 0.27 0.43 0.69 0.27
iJES 0.49 n.a. 0.95 1.05 0.47 0.55 0.72 n.a. 0.33 n.a. n.a. 0.24
stdJES 0.53 2.99 0.45 0.44 0.63 0.56 0.70 0.89 0.24 0.78 0.78 0.20
flavourJES 0.09 0.14 0.03 0.10 0.26 0.40 0.36 0.02 0.11 0.58 0.58 0.12
bJES 0.16 0.33 0.15 0.17 0.07 0.20 0.08 0.71 0.61 0.76 0.49 0.25
MC 0.56 0.36 0.49 0.48 0.63 0.50 0.35 0.64 0.15 0.06 0.28 0.38
Rad 0.06 0.22 0.10 0.28 0.26 0.30 0.45 0.37 0.30 0.58 0.33 0.21
CR 0.21 0.51 0.32 0.28 0.28 0.55 0.32 0.29 0.54 0.13 0.15 0.31
PDF 0.08 0.31 0.19 0.16 0.21 0.30 0.17 0.12 0.07 0.09 0.06 0.09
DetMod < 0.01 <0.01 <0.01 <0.01 0.36 0.50 0.23 0.22 0.24 0.18 0.28 0.10
b-tag 0.03 n.e. 0.10 n.e. 0.10 <0.01 0.81 0.46 0.12 0.09 0.06 0.11
LepPt 0.03 0.27 n.a. n.a. 0.18 0.35 0.04 0.12 0.02 0.14 n.a. 0.02
BGMC 0.12 0.24 n.a. n.a. 0.18 n.a. n.a. 0.14 0.13 0.05 n.a. 0.10
BGData 0.16 0.14 0.56 0.15 0.21 0.20 0.10 n.a. n.a. n.a. 0.13 0.07
Meth 0.05 0.12 0.38 0.21 0.16 0.51 0.13 0.07 0.06 0.40 0.13 0.05
MHI 0.07 0.23 0.08 0.18 0.05 <0.01 0.03 0.01 0.07 0.11 0.06 0.04
Total Syst 0.99 3.13 1.41 1.36 1.25 1.49 1.53 1.50 1.03 1.46 1.23 0.71
Total 1.12 3.69 2.01 1.85 1.50 2.79 1.55 1.63 1.06 1.52 1.41 0.76
Table 3: Uncertainty categories assignment for the input measurements and the result of the world m com-
Luca BrigliadoriUniversity of Bologna
Rencontres de Moriond EWMarch 19, 2014