Critique of Exotica Music, Politics and the Culture ... - Trinketization
Searches for charged Higgs bosons, supersymmetry, and exotica with tau leptons with the ATLAS and...
Transcript of Searches for charged Higgs bosons, supersymmetry, and exotica with tau leptons with the ATLAS and...
Ryan Reece
University of Pennsylvania
September 20, 2012Tau 2012, Nagoya, Japan
Searches for charged Higgs bosons,
supersymmetry, and exotica with
tau leptons with the ATLAS and
CMS detectors at the LHC
on behalf of the ATLAS and CMS collaborations
Ryan Reece (Penn)
Outline
2
• Motivational questions about the SM
• Large Hadron Collider, ATLAS, and CMS
• Charged Higgs searches
• SUSY searches
• Exotics: Leptoquark and Z’ searches
... all with taus.
Ryan Reece (Penn)
Why the Standard Model?
3
• Why the gauge group SU(3)C × SU(2)L × U(1)Y ?
• Why are there 3 generations of quarks and leptons?
• Why are lepton and hadron charges quantized in the
same units? Why the existing hypercharges?
Is it because...
• the gauge group of Nature is actually bigger?
SO(10) → SU(5) × U(1) Georgi-Glashow
SO(10) → SU(4)C × SU(2)L × SU(2)R Pati-Salam
• e.g. Pati-Salam SO(10): QEM = T3L + T3R + 1/2(B − L)
• lifetime of the proton > 1033 years ⇒ if unification happens it must suppress proton decay, e.g. it happens at a high energy scale
QEM = T3L + Y/2
Ryan Reece (Penn)
Gauge coupling unification
4
Using renormalization group equations one can evolve
the electroweak and strong coupling constants, as
measured at LEP.[Amaldi, de Boer, and Furstenau 1991]
SUSY allows
for unification
Ryan Reece (Penn)
GUT Motivations
5
It is interesting that:
• the structure of SM can be embedded in larger groups, and this
could explain the SM hypercharges,
• the SM couplings apparently converge when run to very high
energies.
Motivates the possibility of grand unification.
Avenues of discovery:
• Higgs sector
• supersymmetry
• leptoquarks
• Z’
102 GeV1016 GeV
Ryan Reece (Penn) 10
Datasets
• Analyses reported here use 2-5 fb-1 of the data collected in 2011 at √s = 7 TeV.
• In 2012, ATLAS and CMS each have collected over 14 fb-1 so far at √s = 8 TeV.
[TWiki:CMSPublic/LumiPublicResults, ATLAS-CONF-2012-042]
Number of Vertices
0 5 10 15 20 25 30A
rbitra
ry U
nits
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
> = 15µData 2011, <
> = 29µData 2011, <
> = 32µData 2011, <
ATLAS Preliminary
= 7 TeVs
Ryan Reece (Penn)
High-pT τh reconstruction
11
Isolation
annulus
pile-up
tau
UE
Count # tracks
in core cone
!"#$
"%&'()*'
+"#$
"%&'()*'
,*-./'
0)().(1*
2-*(3.%)45%16
• τh reco seeded by calorimeter jets
• associate tracks in ∆R < 0.2, select 1 or 3
• combine calorimeter and tracking
information in a BDT or likelihood
discriminant, preferring narrow clustering,
hadronic activity[ATLAS-CONF-2011-152, CMS PAS TAU-11-001]
• particle-flow reconstructs constituent 4-vectors
• τh reco seeded by particle-flow hadrons
• Hadron Plus Strip (HPS) algorithm for
counting π0s
• isolation cone for rejecting QCD jets
•Hadronic decays dominantly to 1 or 3 π± and possibly a few additional π0s
•Decay in beam-pipe: cτ ≈ 90 µm
Ryan Reece (Penn)
Charged Higgs
• 2HDM, five spin-0 states:
CP-even H±, CP-even H and h, CP-odd A
• Necessary for MSSM
• tan β = vu/vd
• For tan β > 2, H+→τν
dominant decay
• Produced through top
decays: t→bH+ (for mH+ < mt)
• Look for tt with enhanced τ.
13
f
f′g
g
g
ντ
τ+
H+
W-
t
t
b
b
-
Ryan Reece (Penn)
Charged Higgs: bbqqτh ‘
14[arxiv:1204.2760]
3.2 tt → bbW∓H±→ bb(qq)(τhν)
Emiss
T
0.5 GeV1/2·
√
∑
pT> 13;
[GeV]Tm
0 50 100 150 200 250 300 350 400
Events
/ 2
0 G
eV
0
10
20
30
40
50
60
70
80
) = 5%+
bH→tB(
Data 2011
τTrue
misidτ→Jet
misidτ→e
Multi-jets
SM + uncertainty
= 130 GeV+H
m
-1 Ldt = 4.6 fb∫
= 7 TeVs
ATLAS
+jetsτ
Ryan Reece (Penn)
Charged Higgs
15[arxiv:1204.2760]
[GeV]+H
m
90 100 110 120 130 140 150 160
+ b
H→
t B
-210
-110
1
Observed CLsExpected
σ 1±σ 2±
ATLAS
Data 2011
= 7 TeVs
-1Ldt = 4.6 fb∫
combined
[GeV]+H
m
90 100 110 120 130 140 150 160
βta
n
0
10
20
30
40
50
60
Median expected exclusion
Observed exclusion 95% CL
theoryσObserved +1
theoryσObserved -1
-1Ldt = 4.6 fb∫Data 2011
=7 TeVs maxh
m
ATLAS
Ryan Reece (Penn)
Supersymmetry
18
• Could help explain: observed , ,
, ...
• Especially interesting large parameter space when the is
the Next-to-Lightest-Supersymmetric-Particle ( )
• The possible consistent
extension to the symmetries of 4-D QFT from
. [Haag–Lopuszanski–Sohnius theorem, 1975]
• No observed SUSY partners yet ⇒ SUSY is broken at a high energy scale.
Ryan Reece (Penn)
SUSY signatures
19
• R-parity-conserving: require high-ETmiss and high-HT
HT = ∑pT(jets)
• R-parity-violating: require high-ST
no stable LSP ⇒ less ETmiss discrim.ST = HT + ∑pT(isolated leptons) + ETmiss
• Why add taus?
• B(N2→ ττN1) or B(C1→ τνN1) can be higher than
other flavors when tan β is high.
• Improve stats even without enhancement
• Constrain all couplings λijk
• CMS incorporates taus into their general multi-lepton search.
• ATLAS and CMS perform dedicated searches with taus.
Ni f
f f
N1
~ ~ ~ ~
1
2λijkLiLjEk
N1 ℓ
ℓ ℓ′
ν′′
λ
Ryan Reece (Penn) [arxiv:1206.3949]
Events
0
10
20
30
40
50
60
70
Events
0
10
20
30
40
50
60
70CMS -1 = 4.98 fb
int=7 TeV Ls
Data
Charge Mis-Id
Prompt-Prompt(irreducible)
Nonprompt-Nonprompt
Prompt-Nonprompt
Region 1: Region 2: Region 3: Region 4: Region 5:
> 80 GeVTH > 200 GeVTH > 450 GeVTH > 450 GeVTH > 450 GeVTH
> 120 GeVmiss
TE > 120 GeVmiss
TE > 50 GeVmiss
TE > 120 GeVmiss
TE > 0 GeVmiss
TE
TH
igh p
)µ
/eµ
µ(e
e/
TLow
p
)µ
/eµ
µ(e
e/
TH
igh p
)µ
/eµ
µ(e
e/
TLow
p)
µ/e
µµ
(ee/
TH
igh p
)µ
/eµ
µ(e
e/
TLow
p
)µ
/eµ
µ(e
e/
TH
igh p
)µ
/eµ
µ(e
e/
Tau c
hannels
)ττ/
τµ/
τ(e
TH
igh p
)µ
/eµ
µ(e
e/
(GeV)T
H0 200 400 600 800 1000 1200 1400
(G
eV
)m
iss
TE
0
50
100
150
200
250 µµ
eeµe
τµ
τe
ττ
Regio
n 1
Regio
n 2
Region 4
Region 3
Region 5
CMS -1 = 4.98 fbint.
= 7 TeV, Ls
SUSY: Same-sign dileptons + jets + MET
20
ETmiss
Ryan Reece (Penn)
• Significantly
extends LEP and
CMS 2010 results.
• Gluino masses
500-1000 GeV
excluded.
• Similar result in
opposite-sign
(in back up). (GeV)0m
500 1000 1500 2000 2500
(G
eV
)1/2
m
100
200
300
400
500
600
700
800
900
1000CMS = 7 TeVs, -1 = 4.98 fbint.L
± l~ LEP2
±
1χ∼ LEP2
No EWSB
= L
SP
τ∼
Non-Convergent RGE's
) = 500g~m(
) = 1000g~m(
) = 1500g~m(
) = 2000g~m(
) = 1000
q~m(
) = 1500
q~m(
) = 2000
q~m(
) = 2500
q~m
(
)=10βtan(
= 0 GeV0
A
> 0µ
= 173.2 GeVt
m
Observed Limit (NLO+NLL with uncertainties)
Expected Limit (NLO+NLL)
)-1 = 35 pbint
NLO Obs. Limit (L
SUSY: Same-sign dileptons + jets + MET
21[arxiv:1206.3949]
ETmiss
CMSSM exclusion
Ryan Reece (Penn)
SUSY: multi-lepton
22[arxiv:1106.0933, CMS PAS SUS-11-013, arvix:1204.5341]
) 2 (GeV/cq~
m600 800 1000 1200 1400
)
2
(G
eV
/cg~
m
600
800
1000
1200
140095% C.L. Limits:
CMS observed
σ 1±CMS expected
σ 2±CMS expected
CMS = 7 TeVs, -1 = 35 pbintL
123λ
122λ
• Harder to exclude couplings to 3rd gen. because of tau ID.
• Exclude squark and gluino masses in 1 TeV range in models with
neutralino LSP decaying through RPV.
• Categorize all events with 3-4 leptons, binning in N(lep), N(τh),
ETmiss , HT, Z/no-Z, no-OSSF, ST ⇒ 114 signal regions (GeV)
q~m
1000 1500 2000
(G
eV
)
g~m
1000
1500
2000
95% C.L. CLs Limits
NLO observed
σ1±NLO expected
σ2±NLO expected
-1 = 7 TeV, L = 4.98 fbsCMS
λ123
λ123
λ122
Ryan Reece (Penn)
! +~
! "
! #
χ1
0
G~
Multileptons and Strong Production
CMS multilepton search with 35/pbq~g~
χ1
0
R ±~
χ1
±
!L ±~
q~g~
SUSYcascade
[arvix:1203.6580, arvix:1204.3852]
SUSY: + jets + taus ,
23
ETmiss
Gauge-Mediated Supersymmetry
Breaking (GMSB) motivated
LSP: very light (∼keV) gravitino (G)
~
τ+µ
muon/muon+jet
pµ
T> 18 GeV
pjet
T> 10 GeV
≥1 jet (50 GeV)
—
1 µ (20 GeV)
≥1 med. (20 GeV)
me, µ
T> 100 GeV
meff > 1000 GeV
Event selection
trig:
ST
= HT + ∑pT(!,τ) + ETmiss
∑
ective mass meff =
~
0 200 400 600 800 1000
Events
/ 5
0 G
eV
-110
1
10
210
310
410ATLAS Preliminary
=7 TeVs -1
L dt = 4.7 fb∫Data 2011
Standard ModelMultijetsW+jetsZ+jetsTop
DiBosonsDrell-Yan
= 50TeVΛGMSB - = 40β tan
= 50TeVΛGMSB - = 20β tan
[GeV]effm0 200 400 600 800 1000
Data
/MC
1
2
>
Ryan Reece (Penn)
SUSY: + jets + taus ,
24[arvix:1203.6580, arvix:1204.3852]
ETmiss
• Combine channels:
1τh+!+jets,
1τh+jets, ≥2τh+jets
• Interpret with GMSB
• Λ = “SUSY breaking
scale”
• OPAL:
Λ > 26 TeV
• ATLAS:
Λ > 47-58 TeV [TeV]Λ
10 20 30 40 50 60 70 80 90 100
βta
n
10
20
30
40
50
60=1grav>0, Cµ=3, 5=250 TeV, NmessGMSB: M
1
0χ∼
1τ∼
CoNLSP
Rl~
Theoryexcl.
(2000 G
eV)
g~
(1800 G
eV)
g~
(1600 G
eV)
g~
(1400 G
eV)
g~
(1200 G
eV)
g~
(1000 G
eV)
g~
(800 G
eV)
g~
(600 G
eV)
g~
(400 G
eV)
g~
1
0χ∼ Rl~
Theoryexcl.
(600 G
eV)
g~
(400 G
eV)
g~
(600 G
eV)
g~
(400 G
eV)
g~
1
0χ∼
Theoryexcl.
ATLAS Preliminary
=7 TeVs, -1 dt = 4.7 fb∫
)theorySUSYσ1 ± limit (
SObserved 95% CL
)expσ1 ±Expected limit (
ATLAS 95% CL limit τ 2-12.05fb
)1τ∼OPAL 95% CL (
)R
µ∼OPAL 95% CL (
OPAL 95% CL
Ryan Reece (Penn)
3rd generation leptoquarks
26[CMS PAS EXO-12-002]
(GeV),bτM0 50 100 150 200 250 300 350 400 450
Even
ts
1
10
210
310
= 7 TeVs, -1CMS 4.8 fb
Data
ttbar
W/Z + jets
Other=450 GeV
LQSignal M
Ryan Reece (Penn)
(GeV)1t~
LQ3/M
200 300 400 500 600 700 800
(pb)
σ×
2 b
)τ
→ 1t~B
r(LQ
3/
-410
-310
-210
-110
1
10
210
310=1β with theory unc.,
theoryσ ×2β
=1333
'λ with theory unc.,
theoryσ ×2 b)τ→
1t~
Br(
with theory unc.theory
σ ×2 b)τ→Br(VLQ3
Expected 95% C.L. upper limit
Observed 95% C.L. upper limit
CMS Preliminary
-1 L dt = 4.8 fb∫
3rd generation leptoquarks
27[CMS PAS EXO-12-002]
(GeV)TS
100 200 300 400 500 600 700 800 900 1000
Even
ts
10
20
30
40
50 = 7 TeVs, -1CMS 4.8 fb
Data
ttbarW/Z + jets
+ ll)ττ*(γZ/
VV=450 GeV
LQSignal M
~
~
Ryan Reece (Penn)
) [GeV]missTE, h
τ, hτ(tot
Tm
0 500 1000 1500
Events
/ 5
0 G
eV
-210
-110
1
10
210
310
410
ATLAS Preliminary-1
dt L = 4.6 fb∫ = 7 TeVs
(a)Data 2011
Multijetττ→*γ/Z
ντ→W
Others
ττ→(1250)Z’
Z’ → ττ → τhτh
28[ATLAS-CONF-2012-067]
Ryan Reece (Penn) [ATLAS-CONF-2012-067]
) [GeV]missTE, hτ, µ(tot
Tm
0 500 1000 1500
Events
/ 5
0 G
eV
-210
-110
1
10
210
310
410ATLAS Preliminary
-1dt L = 4.6 fb∫ = 7 TeVs
(b)Data 2011
ττ→*γ/Z
+jetsW
Multijet
µµ→Z
tt
Dibosonsingle top
ττ→(1000)Z’
Z’ → ττ → µτh
29
Ryan Reece (Penn)
• τhτh: 1.25 (1.35) TeV
• µτh: 1.00 (1.05) TeV
• eµ: 0.75 (0.80) TeV
ATLAS Z’ SSM Exclusions: observed (expected) @ 95% CL
Combined limit
30[ATLAS-CONF-2012-067]
[GeV]Z’
m
500 1000 1500
[pb]
Bσ
-310
-210
-110
1
ATLAS Preliminary-1
dt L = 4.7 fb∫ = 7 TeVs
Combined
Expected limit
σ 1±Expected
σ 2±Expected
Observed limit
SSMZ’
(GeV)Z'M400 600 800 1000 1200 1400 1600
) (p
b)
ττ→
BR
(Z'
×Z
')
→(p
pσ
-310
-210
-110
1
10
210-1
= 7 TeV, 4.94 fbs CMS,
Obs. Combined 95% CL
Exp. Combined 95% CL
expected rangeσ1
expected rangeσ2
SSMZ'
ψZ'
1.4 TeV
1.3 TeV 1.4 Te
V1.1 TeV
Ryan Reece (Penn)
Conclusions
31
• The performance of the LHC, and the ATLAS and
CMS experiments have extended many exclusions
for new physics.
• Many searches will be improved with the 2012
dataset and further their reach with increases in
luminosity and energy after the coming long shutdown.
• Unification and supersymmetry remain hidden.
• A new boson is in grasp consistent with the SM
Higgs with a mass of 126 GeV, and taus will play an
important role determining its couplings to fermions!
Ryan Reece (Penn)
General references
• J. Beringer et al. (Particle Data Group), Phys. Rev. D86, 010001 (2012).
• U. Amaldi, W. de Boer, and H. Furstenau. “Comparison of grand unified theories
with electroweak and strong coupling constants measured at LEP.” PLB 260
(1991) 447.
• W. de Boer, C. Sander. “Global Electroweak Fits and Gauge Coupling
Unification.” (2003) [arxiv:0307049]
• S.P. Martin. “A Supersymmetry Primer” [arxiv:9709356]
33
Ryan Reece (Penn)
Charged Higgs
• ATLAS
• “Search for charged Higgs bosons decaying via H± → τ±ν in tt events using pp
collision data at √s = 7 TeV with the ATLAS detector” [arxiv:1204.2760]
• CMS
• “Search for a light charged Higgs boson in top quark decays in pp collisions at
√s = 7 TeV” [arvix:1205.5736]
34
Ryan Reece (Penn)
SUSY
• ATLAS
• “Search for events with large missing transverse momentum, jets, and at least two
tau leptons in 7 TeV proton-proton collision data with the ATLAS detector” [arxiv:
1203.6580]
• “Search for supersymmetry with jets, missing transverse momentum and at least one
hadronically decaying τ lepton in proton-proton collisions at √s = 7 TeV with the
ATLAS detector” [arxiv:1204.3852]
• CMS
• “Search for anomalous production of multilepton events in pp collisions at √s = 7
TeV” [arvix:1204.5341]
• “Search for new physics with same-sign isolated dilepton events with jets and
missing transverse energy” [arxiv:1205.6615]
• “Search for new physics in events with opposite-sign leptons, jets, and missing
transverse energy in pp collisions at √s = 7 TeV” [arxiv:1206.3949]
35
Ryan Reece (Penn)
Exotics
• ATLAS
• “A search for high-mass resonances decaying to τ+τ− in pp collisions at √s = 7
TeV with the ATLAS detector” [ATLAS-CONF-2012-067]
• CMS
• “Search for high-mass resonances decaying into τ-lepton pairs in pp collisions
at √s = 7 TeV” [arvix:1206.1725]
• “Search for pair production of third generation leptoquarks and stops that
decay to a tau and a b quark” [CMS PAS EXO-12-002]
36
Ryan Reece (Penn)
Tau performance
• ATLAS
• “Reconstruction, Energy Calibration, and Identification of Hadronically
Decaying Tau Leptons” [ATLAS-CONF-2011-077]
• “Performance of the Reconstruction and Identification of Hadronic Tau Decays
with ATLAS” [ATLAS-CONF-2011-152]
• “Z → ττ cross section measurement in proton-proton collisions at 7 TeV with
the ATLAS experiment” [ATLAS-CONF-2012-006]
37
Ryan Reece (Penn)
Tau performance• CMS
• “Performance of τ-lepton reconstruction and identification in CMS”
[arvix:1109.6034, CMS PAS TAU-11-001]
• “CMS Strategies for tau reconstruction and identification using particle-flow
techniques” [CMS PAS PFT-08-001]
• “Particle–Flow Event Reconstruction in CMS and Performance for Jets, Taus,
and ETmiss” [CMS PAS PFT-09-001]
• “Commissioning of the Particle-Flow Reconstruction in Minimum-Bias and Jet
Events from pp Collisions at 7 TeV” [CMS PAS PFT-10-002]
• “Commissioning of the particle-flow event reconstruction with leptons from
J/Psi and W decays at 7 TeV” [CMS PAS PFT-10-003]
• “Study of tau reconstruction algorithms using pp collisions data collected at
√s = 7 TeV” [CMS PAS PFT-10-004]
38
Ryan Reece (Penn) 40
ATLAS DetectorMagnets: 5 tonne central solenoid, 2T in inner detector, 4T toroid system
Muon Spectrometer: |η|<2.7, drift-tube chambers
Tracking: |η|<2.5, B=2T, precise tracking and vertexing, Si pixels, strips, and TRT straws, TR electron ID
Electromagnetic Calorimeter: |η|<3.2, 3+1 layers corrugated layers of lead and LAr
Hadronic Calorimeter: |η|<5, Central: iron/scintillator tiles, Forward: copper/tungsten-LAr
Ryan Reece (Penn)
Phenomenology of tau decays
44
τ−
→ e− νe ντ
µ−
νµ ντ
17.8%17.4%
}
leptonic 35.2%
π−
π0ντ
π−
ντ
π− 2π0
ντ
K− (Nπ0) (NK0) ντ
π− 3π0
ντ
25.5%10.9%9.3%1.5%1.0%
1 prong 49.5%
π−
π−
π+
ντ
π−
π−
π+
π0ντ
9.0%4.6%
}
3 prong 15.2%
Ryan Reece (Penn)
pass/f
ail
0
0.05
0.1
W CR: Inclusive, 1p
W CR: OS, 1p
W CR: SS, 1p
Tp
0 50 100 150 200 250
ratio
0
1
2
pass/f
ail
0
0.05
0.1
QCD CR: Inclusive, 1p
QCD CR: OS, 1p
QCD CR: SS, 1p
Tp
0 50 100 150 200 250
ratio
0
1
2
BDT Medium BDT Medium
Observed variance in fake-rates
45
(BDTMedium)
1. Why do quarks and gluons have different tau fake-rates?
2. How does the quark/gluon fraction vary among samples?
• Hypothesis: quarks vs gluons
• Divide the issue into two questions:
ATLAS work in progress ATLAS work in progress
Ryan Reece (Penn)
Jet width for quark/gluons
46
J. Gallicchio, M. Schwartz. “Quark and Gluon Tagging at the LHC”. arXiv:1106.3076.
• !(r) = fraction of jet
energy within ∆R < r.
• Quark jets are more
narrow than gluon jets
of the same energy.
• Tau identification prefers
narrow candidates.
• This is consistent with samples of quark-enriched jets, like
W+jet, having higher fake-rates.
Ryan Reece (Penn)
OS vs SS W+jet
47
q
Wg
q′
(a)
q W
g q′
(b)
q
q′W
g
(c)
• The charge of the quark should correlate with the
reconstructed charge of the tau candidate, therefore (a) and
(b) preferably produce opposite sign W+jet events.
• OS and SS will have different quark/gluon fractions.
Leading order W+jet production:
Ryan Reece (Penn)
Madgraph predicted Quark/Gluon
48
50 100 200 400 800 1600
Q
G
0%
100%
80%
60%
40%
20%
pT Cut on All Jets (GeV)50 100 200 400 800 1600
Q
G
0%
100%
80%
60%
40%
20%
pT Cut on All Jets (GeV)
J. Gallicchio, M. Schwartz. “Pure Samples of Quark and Gluon Jets at the LHC”. arXiv:1104.1175
50 100 200 400 800 16000%
100%
80%
60%
40%
20%
GG
QG
pT Cut on All Jets (GeV)50 100 200 400 800 1600
0%
100%
80%
60%
40%
20%
GGG
QGG
QQG
QQQ
pT Cut on All Jets (GeV)
Ryan Reece (Penn)
Tau identification variables
50
Electrmagnetic radius: REM =
∑∆Ri<0.4
i∈{EM 0−2}EEM
T,i∆Ri
∑∆Ri<0.4
i∈{EM 0−2}EEM
T,i
Track radius: Rtrack =
∑∆Ri<0.4
i pT,i ∆Ri∑∆Ri<0.4
i pT,i
Leading track momentum fraction: ftrack =
ptrackT,1
pτT
Core energy fraction: fcore =
∑∆Ri<0.1
i∈{all}EEM
T,i∑∆Ri<0.4
i∈{all}EEM
T,i
Electromagnetic fraction: fEM =
∑∆Ri<0.4
i∈{EM 0−2}EEM
T,i
∑∆Rj<0.4
j∈{all}EEM
T,j
Cluster mass: mclusters, invariant mass clusters at the EM
energy scale.
Track mass: mtracks, invariant mass of the track system.
Transverse flight path significance: SflightT
Motivation: taus tend to be collimated more than jets, have a
leading track, and often significant neutral pion deposits in the EM
calorimeter.
Ryan Reece (Penn)
[GeV] Tp
0 10 20 30 40 50 60 70 80 90 100
bkgd
ε
-210
-110
1
-1Integrated Luminosity 244 nb / Loose Cuts (Data/MC) / Medium Cuts (Data/MC) / Tight Cuts Data(MC)
PreliminaryATLAS PreliminaryATLAS
EMR
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
candid
ate
s / 0
.01
τN
um
ber
of
0
20
40
60
80
100
120
140
160
180
200
220
310×
ATLAS Preliminary = 7 TeV )sData 2010 (
Pythia QCD JetsττPythia Z->
-1Integrated Luminosity 15.6 nb
First data
51
“Reconstruction of hadronic tau candidates in QCD events at ATLAS with 7 TeV pp collisions”
[ATLAS-CONF-2010-059]
“Tau Reconstruction and Identification Performance in ATLAS”
[ATLAS-CONF-2010-086]
• First comparisons of background distributions and the QCD
fake-rate between data and Monte Carlo.
• Already see that MC over-estimates the jet fake-rate. ⇒kW ≈ 0.5
Ryan Reece (Penn)
Tau discriminants
52
CutspT-parametrized cuts on REM and
Rtrack, and a cut on ftrack.
Projective likelihood
d = ln
(
LS
LB
)
=∑N
i=1 ln
(
pSi(xi)
pBi(xi)
)
Boosted decision trees (BDT)
[GeV]T
p
20 30 40 50 60 70 80 90 100
⟩tr
ack
R⟨
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
Likelihood Score
-20 -15 -10 -5 0 5 10 15 20
Arb
itra
ry U
nits
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18ττ→+Zντ→W
dijet Monte Carlo-1dt L = 23 pb∫2010 dijet data
<60 GeVT
3 prongs 15 GeV<p
ATLAS Preliminary
BDT Score
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1A
rbitra
ry U
nits
0
0.02
0.04
0.06
0.08
0.1
0.12 ττ→+Zντ→W
dijet Monte Carlo-1dt L = 23 pb∫2010 dijet data
<60 GeVT
3 prongs 15 GeV<p
ATLAS Preliminary
ATLAS work in progress
Ryan Reece (Penn)
Maturing of discriminants
53
Signal Efficiency
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Inve
rse
Backgro
und E
ffic
ien
cy
1
10
210
310
Cuts
BDT
Likelihood
>20GeVT
1-Prong p
ATLAS Preliminary
[GeV]T
p
20 30 40 50 60 70 80 90 100
⟩E
MR⟨
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
(a) R cut curves for 1-prong
• Cuts are pt-parametrized to account for the Lorentz collimation
of boosted taus.
• Experience grows with LLH and BDT discriminants, which
become the preferred discriminants in 2011.
“Reconstruction, Energy Calibration, and Identification of Hadronically Decaying Tau Leptons
in the ATLAS Experiment” [ATLAS-CONF-2011-077, ATL-PHYS-INT-2011-068]
ATLAS work in progress
Ryan Reece (Penn)
Seeing first hadronic taus
54
Number of tracks
0 1 2 3 4 5 6 7 8 9 10
Num
ber
of
events
0
500
1000
1500
2000
2500
0 1 2 3 4 5 6 7 8 9 100
500
1000
1500
2000
2500
0 1 2 3 4 5 6 7 8 9 100
500
1000
1500
2000
2500
0 1 2 3 4 5 6 7 8 9 100
500
1000
1500
2000
2500
0 1 2 3 4 5 6 7 8 9 100
500
1000
1500
2000
2500
0 1 2 3 4 5 6 7 8 9 100
500
1000
1500
2000
2500
0 1 2 3 4 5 6 7 8 9 100
500
1000
1500
2000
2500
0 1 2 3 4 5 6 7 8 9 100
500
1000
1500
2000
2500 = 7 TeV)sData 2010 (
τν hτ →W
EW background
QCD background (B)
ATLAS Preliminary
-1 L dt = 34 pb∫
EMR
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14
Num
ber
of
events
/ 0
.01
0
100
200
300
400
500
600
700
0 0.02 0.04 0.06 0.08 0.1 0.12 0.140
100
200
300
400
500
600
700
0 0.02 0.04 0.06 0.08 0.1 0.12 0.140
100
200
300
400
500
600
700
0 0.02 0.04 0.06 0.08 0.1 0.12 0.140
100
200
300
400
500
600
700
0 0.02 0.04 0.06 0.08 0.1 0.12 0.140
100
200
300
400
500
600
700
0 0.02 0.04 0.06 0.08 0.1 0.12 0.140
100
200
300
400
500
600
700
0 0.02 0.04 0.06 0.08 0.1 0.12 0.140
100
200
300
400
500
600
700
0 0.02 0.04 0.06 0.08 0.1 0.12 0.140
100
200
300
400
500
600
700 = 7 TeV)sData 2010 (
τν hτ →W
EW background
QCD background (BD)
ATLAS Preliminary
-1 L dt = 34 pb∫
• Nov 2010: Observation of W→τhν [ATLAS-CONF-2010-097]
• Feb 2011: Observation of Z→τhτl [ATLAS-CONF-2011-010]
Ryan Reece (Penn)
W→τν cross section
55
σ(W → τν) = 11.1 ± 0.3(stat.)± 1.7(sys.)± 0.4(lumi.) nb
σtheory = 10.46 ± 0.52 nb at NNLO
) [nb]lν l →(W σ
6 7 8 9 10 11 12 13 14 15 16
τν τ →W ATLAS
eν e →W ATLAS
µν µ →W ATLAS
= 7 TeV)sData 2010 (
Stat uncertainty
Stat ⊕Sys
Lumi⊕ Stat ⊕Sys
Prediction (NNLO)
Theory uncertainty
= 7 TeV)sData 2010 (
Stat uncertainty
Stat ⊕Sys
Lumi⊕ Stat ⊕Sys
Prediction (NNLO)
Theory uncertainty
ATLAS Preliminary
) [nb]lν l →(W σ
6 7 8 9 10 11 12 13 14 15 16
τν τ →W ATLAS
eν e →W ATLAS
µν µ →W ATLAS
Dominant systematics
τh efficiency 10.3%
τh energy scale 8.0%
τh + MET trigger
efficiency 7.0%
luminosity 3.4%
acceptance 2.3%
“Measurement of the W→τν cross section in pp collisions at sqrt(s)= 7 TeV with the ATLAS experiment”
[arXiv:1108.4101]
Ryan Reece (Penn)
Z→ττ cross section
56
σcombined = 0.97 ± 0.07(stat.)± 0.07(sys.)± 0.03(lumi.) nb
σtheory = 0.96 ± 0.05 nb at NNLO
<116 GeV) [nb]inv
ll, 66<m→(Z σ0.6 0.8 1 1.2 1.4 1.6
-136pb
combinedττ →Z
-133-36pb
µµ ee/→Z
hτ µτ
hτ eτ
µτ eτ
µτ µτ
Stat
Stat ⊕Syst
Lumi⊕ Stat ⊕Syst
Theory (NNLO)
Stat
Stat ⊕Syst
Lumi⊕ Stat ⊕Syst
Theory (NNLO)
ATLAS Preliminary
<116 GeV) [nb]inv
ll, 66<m→(Z σ0.6 0.8 1 1.2 1.4 1.6
-136pb
combinedττ →Z
-133-36pb
µµ ee/→Z
hτ µτ
hτ eτ
µτ eτ
µτ µτ
Dominant systematics
τh energy scale 11%
τh efficiency 8.6%
µ efficiency 8.6%
e efficiency 3-10%
acceptance 3%
luminosity 3.4%
“Measurement of the Z→ττ cross section in pp collisions at sqrt(s)= 7 TeV with the ATLAS detector”
[arXiv:1108.2016]
Ryan Reece (Penn)
First systematic recommendations
57
• Systematic uncertainties estimated with dedicated Monte Carlo
with shifts in UE, hadronization, and detector-related effects.
• The efficiency measurement has been superseded with data-
driven measurements from Z→ττ and W→τν tag-and-probe.[ATLAS-CONF-2011-077]
20 40 60 80 100
Ta
u e
ffic
ien
cy
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Looser Cuts working point
1 prong
ATLAS Preliminary
Nominal
Detector Material
Hadronic ShowerUnderlying Event
Noise threshold
Total sys. error
of truth tau [GeV]T
Visible p20 40 60 80 100
Ratio
0.8
0.9
1
1.1
1.2
Ta
u e
ffic
ien
cy 1
[GeV]gen
Tp
10 20 30 40 50 60 70
)gen
T/p
TE
S
T(p
∆
-0.04
-0.02
0
0.02
0.04
0.06
0.08 Underlying eventHadronic showerDetector materialNoise thresholdEM ScaleNon-closureTotal sys. error
Underlying eventHadronic showerDetector materialNoise thresholdEM ScaleNon-closureTotal sys. error
Underlying eventHadronic showerDetector materialNoise thresholdEM ScaleNon-closureTotal sys. error
Underlying eventHadronic showerDetector materialNoise thresholdEM ScaleNon-closureTotal sys. error
Underlying eventHadronic showerDetector materialNoise thresholdEM ScaleNon-closureTotal sys. error
Underlying eventHadronic showerDetector materialNoise thresholdEM ScaleNon-closureTotal sys. error
Underlying eventHadronic showerDetector materialNoise thresholdEM ScaleNon-closureTotal sys. error
ATLAS Preliminary 1.3≤| η1 prong: |
)
Ryan Reece (Penn)
W→τν tag-and-probe
58
Number of tracks0 2 4 6 8 10 12 14 16
Events
0
200
400
600
800
1000
1200
ATLAS Preliminary
Before tau ID
)-1Data 2010 (34 pb
)ντ → (Wτ
)ν e→e (W
Jet background
Number of tracks0 2 4 6 8 10 12 14 16
Events
0
100
200
300
400
500
600
700
800
ATLAS Preliminary
After tau ID
(Looser Cuts working point)
)-1Data 2010 (34 pb
)ντ → (Wτ
)ν e→e (W
Jet background
Tag jet + MET events, probe for tau.
Tau identification efficiency scale factor0.5 1 1.5
0
7
Boosted decision trees 0.13(sys)± 0.07(stat) ±0.94 (Cross section)
Likelihood 0.16(sys)± 0.04(stat) ±1.02 (Cross section)
Cuts 0.13(sys)± 0.05(stat) ±1.00 (Cross section)
Boosted decision trees 0.05(sys)± 0.06(stat) ±1.05 (Tag&Probe)
Likelihood 0.05(sys)± 0.09(stat) ±1.02 (Tag&Probe)
Cuts 0.04(sys)± 0.06(stat) ±1.04 (Tag&Probe)
Looser identification working point-1L = 34 pbATLAS Preliminary
apply
ID
“Measurement of hadronic tau identification efficiency with W→τν events”
[ATLAS-CONF-2011-093]
Ryan Reece (Penn)
CMS Particle Flow
60
!"#$%&'()*)+,&-
!"#$%!&'()&*+
!"#$%!&'()&*+
,"#$%!&'()&*+
,"#$%!&'()&*+
.(/01'&$()+2'3$14$1&,.2
/+()+2'3$.(/01'&
1&,.2'&$5)1
0''$!"#$%6#$%789:;9;;<
• Matches track to clusters to form charged and neutral PF objects.
• PF objects are used as input for all CMS tau reconstruction.
Ryan Reece (Penn)
CMS: Hadron Plus Strip (HPS)
61
Discrimination with calorimeter based isolation ∆R < 0.5.
[CMS PAS TAU-11-001]
Ryan Reece (Penn)
CMS: Tau Neural Classifier (TaNC)
62
• Uses a shrinking core-cone:
• ∆R(photons) < 0.15 for photons
• ∆R(charged) < (5 GeV)/ET for charged hadrons
• ∆R(charged) < ∆R(isolation) < 0.5
• Immediately discarded if the candidate doesn’t match
an expected tau decay mode.
• Dedicated Neural-net classifier for each decay mode
Decay mode Resonance Mass (MeV/c2) Branching fraction (%)
τ− → h−ντ 11.6%
τ− → h−π0ντ ρ− 770 26.0%
τ− → h−π0π0ντ a
−
1 1200 9.5%
τ− → h−
h+
h−ντ a
−
1 1200 9.8%
τ− → h−
h+
h−π0ντ 4.8%
[CMS PAS TAU-11-001]
Ryan Reece (Penn)
CMS Performance
63
(GeV/c)hτ
Tgenerated p
0 50 100
effic
ien
cy
τexpecte
d
0
0.2
0.4
0.6
0.8
1
HPS loose
HPS mediumHPS tight
= 7 TeVsCMS Simulation,
(GeV/c)hτ
Tgenerated p
0 50 100
effic
ien
cy
τexpecte
d
0
0.2
0.4
0.6
0.8
1
TaNC loose
TaNC mediumTaNC tight
= 7 TeVsCMS Simulation,
[CMS PAS TAU-11-001]
0 50 100 150 200
mis
identification r
ate
for
jets
τ
-310
-210
Dataνµ→W
Simulationνµ→W
QCD Data
QCD Simulation
DataµQCD
SimulationµQCD
HPS loose
-1 = 7 TeV, 36 pbsCMS,
(GeV/c)T
jet p0 50 100 150 200
Sim
ula
tion
Data
-Sim
.
-0.2
0
0.20 50 100 150 200
mis
identification r
ate
for
jets
τ
-310
-210
Dataνµ→W
Simulationνµ→W
QCD Data
QCD Simulation
DataµQCD
SimulationµQCD
TaNC loose
-1 = 7 TeV, 36 pbsCMS,
(GeV/c)T
jet p0 50 100 150 200
Sim
ula
tion
Data
-Sim
.
-0.2
0
0.2
• Not trivial to
compare ATLAS and
CMS tau
performance because
we bin fake-rates in
N(track) instead of
categorizing the
decay mode.
Ryan Reece (Penn)
CMS decay mode ID
64
0.85 0.16 0.05
0.13 0.83 0.04
0.02 0.01 0.91
decay modeτgenerated
π )0π(0ππ πππ
decay m
ode
τre
constr
ucte
d
π
0ππ
πππ
= 7 TeV sCMS Simulation,
decay modeτreconstructed
π 0ππ πππ
rela
tive y
ield
0
0.2
0.4
0.6
0.8
Dataτ τ →Z
W+jets
/ewkt tQCD
-1 = 7 TeV, 36 pbsCMS,
[CMS PAS TAU-11-001]
Ryan Reece (Penn)
Calorimeter granularity
65
• B = 3.8 T
• ∆η ×∆! = 0.0174×0.0174
• R = 0.5 anti-kT PF-jets
• B = 2.0 T
• ∆η ×∆! = 0.025×0.0245
• R = 0.4 anti-kT topo-jets
∆ϕ = 0.0245
∆η = 0.02537.5mm/8 = 4.69 mm ∆η = 0.0031
∆ϕ=0.0245x4 36.8mmx4 =147.3mm
Trigger Tower
TriggerTower∆ϕ = 0.0982
∆η = 0.1
16X0
4.3X0
2X0
1500
mm
470
mm
η
ϕ
η = 0
Strip cells in Layer 1
Square cells in Layer 2
1.7X0
Cells in Layer 3 ∆ϕ× ∆η = 0.0245× 0.05
ATLAS
CMS
ATLAS Barrel EM Calorimeter
Granularity could fundamentally limit our capacity to
reconstruct sub-structure / π0s.
0.0250.0174
Ryan Reece (Penn)
Charged Higgs: bblτh ‘
67[arxiv:1204.2760]
[GeV]miss
TE
0 50 100 150 200 250 300
Events
/ 2
0 G
eV
0
50
100
150
200
250
300
350
400
) = 5%+
bH→tB(
Data 2011
τTrue
misidτ→Jet
misidτ→e
Misid’ed lepton
SM + uncertainty
= 130 GeV+H
m
-1 Ldt = 4.6 fb∫
= 7 TeVs
µ+τ
ATLAS
3.1 tt → bbW∓
H±→ bb(l∓ν)(τhν)
Ryan Reece (Penn)
(GeV)0m
500 1000 1500 2000 2500
(G
eV
)1/2
m
100
200
300
400
500
600
700
800
900
1000CMS = 7 TeVs, -1 = 4.98 fbint.L
± l~ LEP2
±
1χ∼ LEP2
No EWSB
= L
SP
τ∼
Non-Convergent RGE's
) = 500g~m(
) = 1000g~m(
) = 1500g~m(
) = 2000g~m(
) = 1000
q~m(
) = 1500
q~m(
) = 2000
q~m(
) = 2500
q~m
(
)=10βtan(
= 0 GeV0
A
> 0µ
= 173.2 GeVt
m
Observed Limit (NLO+NLL with uncertainties)
Expected Limit (NLO+NLL)
)-1 = 35 pbint
NLO Obs. Limit (L
ATLAS: Same-sign dileptons + jets + MET
71[ATLAS-CONF-2012-105, arxiv:1206.3949]
ETmiss
• Select SS ee, µµ, eµ + 4 jets + ETmiss > 150 GeV
• ATLAS result does not use hadronic taus.
[GeV]0
m
500 1000 1500 2000 2500 3000 3500
[G
eV
]1/2
m
100
200
300
400
500
600
700
(600 G
eV
)
q~
(1000 G
eV
)
q~
(1400 G
eV
)
q~
(600 GeV)g~
(1000 GeV)g~
(1400 GeV)g~
>0µ= 0, 0
= 10, AβMSUGRA/CMSSM: tan
=8 TeVs, -1
L dt = 5.8 fb∫T
miss 4 jets + E≥SS dilepton +
ATLAS )theory
SUSYσ1 ±Observed limit (
)expσ1 ±Expected limit (
SAll limits at 95% CL
1
±χ∼LEP2
Non-convergent RGE
No EW SB
=7 TeVs -1SS dilepton, 2 fb
Preliminary
Ryan Reece (Penn)
SUSY: + jets + taus ,
72
ETmiss
Event selection
≥1τh channel
[ arvix:1204.3852]
[GeV]eff
m
0 200 400 600 800 1000 1200 1400 1600
Events
/ 1
00 G
eV
10
20
30
40
50
60Total SM
multijets
Z + Jets
W + Jets
top
Data 2011
GMSB(40,30)
GMSB(30,20)
ATLAS-1
Ldt = 2.05 fb∫
Ryan Reece (Penn)
SUSY: multi-lepton
74[arxiv:1106.0933, CMS PAS SUS-11-013, arvix:1204.5341]
(GeV) 1
±χ∼
m500 600 700 800 900
(G
eV
)
g~m
1500
1600
1700
1800
1900
2000
95% C.L. CLs Limits
NLO observed
σ1±NLO expected
σ2±NLO expected
-1 = 7 TeV, L = 4.98 fbsCMS
• Deviation in obs. limit from slight excess in category: 4lep(1τh), ETmiss > 50 GeV, HT < 200 GeV, no Z
Expect 0.59 ± 0.17 and observed 3 events.
Ryan Reece (Penn) [arxiv:1206.3949]
SUSY: Opposite-sign dileptons + jets + MET
75
ETmiss
Events
0
10
20
30
40
50
60
70
Events
0
10
20
30
40
50
60
70CMS -1 = 4.98 fb
int=7 TeV Ls
Data
Charge Mis-Id
Prompt-Prompt(irreducible)
Nonprompt-Nonprompt
Prompt-Nonprompt
Region 1: Region 2: Region 3: Region 4: Region 5:
> 80 GeVTH > 200 GeVTH > 450 GeVTH > 450 GeVTH > 450 GeVTH
> 120 GeVmiss
TE > 120 GeVmiss
TE > 50 GeVmiss
TE > 120 GeVmiss
TE > 0 GeVmiss
TE
TH
igh p
)µ
/eµ
µ(e
e/
TLow
p
)µ
/eµ
µ(e
e/
TH
igh p
)µ
/eµ
µ(e
e/
TLow
p)
µ/e
µµ
(ee/
TH
igh p
)µ
/eµ
µ(e
e/
TLow
p
)µ
/eµ
µ(e
e/
TH
igh p
)µ
/eµ
µ(e
e/
Tau c
hannels
)ττ/
τµ/
τ(e
TH
igh p
)µ
/eµ
µ(e
e/
Ryan Reece (Penn)
(GeV)0m500 1000 1500 2000 2500 3000
(G
eV
)1/2
m
100
150
200
250
300
350
400
450
500
550
600
± l~ LEP2
±
1χ∼ LEP2
No EW
SB
= L
SP
τ∼
Non-Conve
rgent R
GE's) = 500g~m(
) = 1000g~m(
) = 1000
q~m(
) = 1
500
q ~m
(
) = 2
000
q ~m
(
) = 2
500
q ~m
(
)=10βtan( = 0 GeV
0A
> 0µ = 173.2 GeVtm
observed
theory)σ1±observed (
expected
stat)σ1±expected (
observed 2010
-1L dt 4.98 fb∫ = 7 TeV, sCMS
LM1
LM3
LM6
LM13
SUSY: Opposite-sign dileptons + jets + MET
76[arxiv:1206.3949]
ETmiss
• Similar methods
to the same-sign
search
• Slightly less
sensitive
Ryan Reece (Penn)
SUSY: + jets + taus ,
77[CMS PAS SUS-12-004]
ETmiss
[GeV]0m
300 400 500 600 700 800 900 1000
[GeV
]1/
2m
100
200
300
400
500
600
700
800
900
1000
± l~ LEP2
±
1χ∼ LEP2
= L
SP
τ∼
) = 500g~m(
) = 1000g~m(
) = 1500g~m(
) = 1000
q~m(
) = 1500q~m(
Obs. LimitExp. Limit
σ 1±Exp. (theory)σ 1±Obs.
)=40βtan(
= -500 GeV0A
> 0µ
= 173.2 GeVtm
(b) CMS Preliminary
= 7 TeVs, -1
4.98 fb
Ryan Reece (Penn)
Data-driven multijet
79
Norm
aliz
ed e
vent
rate
/ 1
0 G
eV
0
0.05
0.1
0.15
0.2
0.25
0.3
ATLAS Preliminary-1
dt L = 4.7 fb! = 7 TeVs
Opposite Charge (OC)
Same Charge (SC)
) [GeV]missTE, h", h"(TM
200 250 300 350
OC
/ S
C
0.51
1.5
Fit same-sign (SS) data with dijet function:
Events
/ 4
0 G
eV
-110
1
10
210
310
410ATLAS Preliminary
-1dt L = 4.7 fb! = 7 TeVs
Data 2011Multijet
""#*$/Z
%"#W
Others""#(1250)Z’
) [GeV]missTE, h", h"(TM
500 1000 1500obs.
/ exp.
0.51
1.5
• OS/SS shapes agree well
• normalize in OS sideband with 200 < MT < 250 GeV[ATLAS-CONF-2012-067]
Ryan Reece (Penn)
) [GeV]µ(Tp
0 50 100 150
Muon I
sola
tion F
ake F
acto
r
0
0.05
0.1
ATLAS Preliminary-1
dt L = 4.7 fb∫ = 7 TeVs
Multijet
Inclusive
|<1.37η|
|<1.52η1.37<|
|<2.5η1.52<|
Multijet background estimation
80
Multijet control region
• In the control region, divide leptons
into pass and fail isolation.
• Define fake factor:
• Predict the number of multijet events:
fµ–iso(pT, η) ≡Npass µ–iso(pT, η)
Nfail µ–iso(pT, η)
∣
∣
∣
∣
∣
∣
multijet–CR
Nmultijet(pT, η, x) = fµ–iso(pT, η) · Nfail µ–iso
multijet(pT, η, x) .
This assumes that the ratio of the number of isolated muons to the number of non-isolated muons inx) = fµ–iso(pT, η) ·
(
Nfail µ–iso
data(pT, η, x) − N
fail µ–iso
MC(pT, η, x)
)
.
The shape of the multijet background in any kinematic variable, x M , is modeled from
Ryan Reece (Penn) ) [GeV]hτ(
Tp
0 50 100 150 200 250
Ta
u I
D F
ake F
acto
r
0
0.05
0.1
0.15
0.2
0.25ATLAS Preliminary
-1dt L = 4.7 fb∫ = 7 TeVs
)+jetsνµ→(W
Inclusive
|<0.8η|
|<1.37η0.8<|
|<2.47η1.52<|
Events
/ 1
0 G
eV
0
200
400
600
800
1000
1200
1400
1600
1800
2000ATLAS Preliminary
-1dt L = 4.7 fb∫ = 7 TeVs
Data 2011ττ→*γ/Z
+jetsW
Multijetµµ→Z
tt
Dibosonsingle top
) [GeV]missTE, µ(Tm
0 50 100 150 200obs.
/ exp.
0.51
1.5
W+jet background estimation
81
W+jet control region
• In a W+jet control region, divide tau candidates into
pass and fail identification.
• Define fake factor:
• Predict the number of W/Z+jet events:
fτ(pT, η) ≡Npass τ−ID(pT, η)
Nfail τ−ID(pT, η)
∣
∣
∣
∣
∣
∣
W–CR
NW/Z+jet(pT, η, x) = fτ(pT, η) · Nfail τ−IDW/Z+jet (pT, η, x) ,
x) = fτ(pT, η) ·(
Nfail τ−IDdata (pT, η, x) − Nfail τ−IDmultijet (pT, η, x) − N
fail τ−IDMC (pT, η, x)
)
.
Ryan Reece (Penn)
Double fake factor procedure
82
!"#$%&
!"#$%&#'()*+(
,-.#&%$.&#$-.
!"#$%&#'
/01'#"
,-.#&%2
$.&#$-.
!"#$%&#'()*+(
,-.#&%$.&#$-.
'(#&"#")&(*+,-(./*01$.%2"$%."1$
'(#&"#")&(*+,-(./*01$.%2"$%."1$'1$3$(#&"#")&(*456*01$.%2"$%."1$
7%//*.%8*96
7%//*&(:.1$*"/1
!"#$%&"'%()
7%//*&(:.1$*"/1
7%//*.%8*96
!"#$%$*+&,-%#.,
!"#$%&"'%()
!"#$%$*+&,-%#.,
• The multijet contamination is estimated from the rate of non-isolated leptons, in
both the signal region that passes tau ID, and the sample that fails.
• Then, the corrected number of tau candidates failing ID are weighted to predicted
the W+jet background.
• This way, the corrections are small at each step.
Ryan Reece (Penn)
) [GeV]miss
TE, µ, e(tot
Tm
0 500 1000 1500
Events
/ 5
0 G
eV
-210
-110
1
10
210
310
410ATLAS Preliminary
-1dt L = 4.6 fb∫ = 7 TeVs
(d)Data 2011
Dibosonττ→*γ/Z
tt+jetsW
µµ→Zττ→(750)Z’
Z’ → ττ → eµ overview
83
Event selection
Dominant systematics
Ryan Reece (Penn)
Z’→ττ
85[arvix:1206.1725]
) (GeV)T
miss,Ehτ,µτ M(
0 200 400 600 800 1000
Events
/ 2
0 G
eV
-210
-110
1
10
210
310-1
= 7 TeV, 4.94 fbs(c) CMS, Data
hτµτ →(750) SSMZ'
τ τ →Z µ µ →Z
ttW+jets
WWQCD
miss
) (GeV)T
miss,Ehτ,
hτ M(
0 200 400 600 800 1000
Events
/ 2
0 G
eV
-210
-110
1
10
210
310-1
= 7 TeV, 4.94 fbs(d) CMS, Data
hτ
hτ →(750) SSMZ'
τ τ →Z
ee→Z W+jets
QCD