Higgs Physics at the Tevatron
Aidan Robson
University of Glasgow
Warwick, 9 December 2010
 Motivation
 Standard Model Basis
and indirect searches
 Higgs searches:
Low mass / high mass channels
 Tevatron Combination
 Limit-setting
 Analysis techniques
 Analysis stability
 Improving sensitivity
 Outlook and LHC
Higgs Physics at the Tevatron
2
QED
ψ: electron field
A: photon field
Higgs Physics at the Tevatron
3
Electroweak
(B and W mix to give physical W±,Z,photon fields)
Higgs Physics at the Tevatron
4
Electroweak
(B and W mix to give physical W±,Z,photon fields)
Higgs Physics at the Tevatron
5
Electroweak
Higgs Physics at the Tevatron
6
WW scattering
W+
W+
Z/g
W–
W–
W+
H
W+
W+
W+
W+
W–
W–
W–
W+
required to
cancel highenergy
behaviour
Z/g
W–
W+
W+
H
W–
W–
W–
Higgs Physics at the Tevatron
W–
7
p
g
H
p
g
Higgs Physics at the Tevatron
8
Tevatron
h = 0.6
h = 1.0
h = 2.0
muon
chambers

D0
CDF
pre-radiator
2
shower max
hadronic cal
had
cal
EM cal
solenoid
1
0
tracker
0
silicon
1
h=1
E
M
cal
2
had cal
h=2
h=3
3 m
Fibre tracker to |h|<1.8
Calorimeter to |h|<4
Muon system to |h|<2
Drift chamber to |h|<1
Further tracking from Si
Calorimeter to |h|<3
Muon system to |h|<1.5
Higgs Physics at the Tevatron
9
Susy
Higgs
dibosons
top
quark
W/Z
Dms=17.77 ±
0.10 ± 0.07
bottom
quark
Jets
Higgs Physics at the Tevatron
10
mW
UK!
mW:
CDF: mW = 80413 ± 48 MeV/c2
D0: mW = 80402 ± 43 MeV/c2
Tev: mW = 80420 ± 31 MeV/c2 (includes Run 1)
LEP: mW = 80376 ± 33 MeV/c2
Heading to CDF 25MeV/c2 measurement
CDF
DmZ (stat)
published (200/pb)
43 MeV
expected (2.3/fb)
13 MeV
Higgs Physics at the Tevatron
11
mt
xobs:
px
py
pz
px
py
pz
lep1
Matrix element-based top mass measurement
Lepton+jets with 4.8fb-1
NN for background discrimination
Likelihood fit over variables sensitive to top mass
Simultaneous constraint of jet energy scale using W in lepton+jets
More precise than CDF 2009!
Expect 1GeV precision achievable
Higgs Physics at the Tevatron
E x , Ey
etc.
jet1 (true values y)
ET model
lepton energy resn
mt =172.8 ± 1.3total GeV
(0.7stat 0.6JES 0.8sys)
12
Indirect constraints
e+
e–
Z
Z
H
mH>114GeV
b
b
mH<154GeV
estimated
final precision
Higgs Physics at the Tevatron
13
Single top
d
u
l
l
W
W
b
t
g
W
Single top observed 2009.
b
b
b
t
n
W
d
s-channel
b
t-channel cross section [pb]
t-channel
n
u
s-channel cross sectionHiggs
[pb]Physics at the Tevatron
14
Dibosons
W/Z
q
W/Z/g
q’
W/Z/g
events
Wg Zg WW tt WZ t ZZ
g
Z
h3, ZZg
g
non-SM
Z
SM
photon ET (GeV)
g
Z
H→
WW
Higgs Physics non-SM
at the Tevatron
|h3| < 0.037, |h4| < 0.0017
@95%CL (L=1.2TeV)
h3, Zgg
15
Dibosons
q
W
q
UK!
Z
W
q’
Z/g
q’
Z
ZZ seen in 4 lepton at 5.7σ
All now observed!
Wg Zg WW tt WZ t ZZ
σ(pp → WZ)
= (4.1 ± 0.7) pb
Higgs Physics at the Tevatron
H→
WW
σ(pp → ZZ )
= (1.7 +1.2-0.7 (stat)
± 0.2 (syst)) pb
16
SM Higgs searches
t,b
g
Br
H
g
q
W/Z
fb
W/Z
ggH
s/
H
q’
qqWH
q
qqqqH
qqZH
q’
mH/GeV
Higgs Physics at the Tevatron
W/Z
W/Z
H
17
ZH → nnbb
UK!
n
q
cut at 0.6 removes 95% of the QCD
background, 65% of the non-QCD
background, and keeps 70% of signal
Z
q’
Z
H
n
b
b
most difficult final
state among most
sensitive channels
Higgs Physics at the Tevatron
18
ZH → nnbb
UK!
EWK and jets control regions
Higgs Physics at the Tevatron
19
Limit setting
Higgs signal x 10
H1=SM+Higgs (of mass mH)
H0=SM only
X
X = some observable
 Construct test statistic Q = P(data|H1)/P(data|H0)
–2lnQ = c2(data|H1) – c2(data|H0) ,
marginalized over nuisance params except s H
0
 Find 95th percentile of resulting s H distribution
– this is 95% CL upper limit.
 Repeat for pseudoexperiments drawn from
expected distributions to build up expected
outcomes
 Median of expected outcomes is “expected limit”
95%
rescale
1
2
sH (pb)
0
2
sH/sSM
Median = expected limit
Expected outcomes
 When computed with collider data this is the
“observed limit”
95%
PDF
signal
separation
Background
events
background
suppression
95% CL Limit/SM
Higgs Physics at the Tevatron
20
Limit setting (2)
mH=160
95% CL Limit / SM
median
Repeat for different values
of mH  build up exclusion plot
Higgs Physics at the Tevatron
1s
2s
illustrative
mH / GeV
21
What have we found?
95% CL Limit / SM
Deficit
Excess
illustrative
expected limit
observed limit
mH / GeV
Higgs Physics at the Tevatron
illustrative
expected limit
observed limit
mH / GeV
22
ZH → nnbb
Exp 4.6xSM
Obs 3.7xSM
UK!
Observe VZ with Z→bb?
PRL 104 071801 (2010)
Higgs Physics at the Tevatron
23
WH → l nbb
Key issue: estimating W+bb background
Shape from MC with normalization from data control regions
Matrix element analysis
q
W
q’
W
H
l
n
b
b
Results at mH = 115GeV: 95%CL Limits/SM
Higgs
Exp.
Obs.
Events
Limit
Limit
19.7
3.5
3.6
World’s most sensitive
low-mass Higgs search
Higgs Physics at the Tevatron
24
ZH → ll bb
q
Z
q’
Z
H
l
l
b
b
Results at mH = 115GeV: 95%CL Limits/SM
Higgs Physics at the Tevatron
Higgs
Exp.
Obs.
Events
Limit
Limit
3
5.48
5.99
25
H → tt
lepton
opposite
hadronic tau
charge
(1&3prong)
2 jets
tlepthad: Br 45%
3 ANN: sig vs:
Ztt, tt, QCD
combined for fit
0
100
MT(lep,ET) /
GeV/c2
Alpgen Z+jets
Higgs Physics at the Tevatron
26
H → gg
g
W
W
H
g
W
Results at mH = 115GeV: 95%CL Limits/SM
Exp.
Obs.
Limit
Limit
20.8
24.6
Higgs Physics at the Tevatron
27/54
WHlnbb(NN)
ZHllbb(NN)
METbb(NN)
WH(l)nbb: ZHnnbb
WHlnbb(ME)
Tag
Lepton
Jet
Tag
Leptons
Tag
SecStx+SecVtx
Tight
2
SecStx+SecVtx
High S/B
SecVtx+JP
Tight
2
SecVtx+JP
High S/B
Tag
SecVtx+JP
SecVtx+Nntagger
Tight
2
SecVtx Single
High S/B
SecStx+SecVtx
SecVtx Single
SecVtx Single
Tight
2
SecStx+SecVtx
Low S/B
SecStx+SecVtx
IsoTrk
2
SecVtx+JP
Low S/B
SecVtx+JP
IsoTrk
2
SecVtx Single
Low S/B
SecVtx+Nntagger
IsoTrk
2
SecVtx Single
IsoTrk
2
SecStx+SecVtx
Forward
2
LooseSecVtx+JP
Forward
2
SecVtx+Nntagger
Forward
2
SecVtx Single
Forward
3
SecStx+SecVtx
Tight
3
LooseSecVtx+JP
Tight
3
SecVtx Single
Tight
3
SecStx+SecVtx
Extend-
3
LooseSecVtx+JP
Extend-
3
SecVtx Single
Extend-
3
SecStx+SecVtx
SecVtx+JP
SecVtx Single
SecStx+SecVtx
SecVtx+JP
SecVtx Single
Many contributions
W/ZHjjbb(NN)
Tag
SecStx+SecVtx
Htt
Hgg
Higgs Physics at the Tevatron
28
Low mass Higgs searches
Higgs Physics at the Tevatron
29
Low mass Higgs searches
Higgs Physics at the Tevatron
30
H→WW
l
g
W
t,b
H
g
variables describing
event topology
kinematics of
two leptons
DY
102
e
DY
102
ETHiggs Physics at the Tevatron
W
n
l
n
Wj
Wg
tt
WZ
ZZ
DY
WW
HWWx10
Data
ee
ET
31
H→WW
4.8fb–1
0 jet WW
4.8fb–1
S:13
4.8fb–1
UK!
1 jet
>1 jet
S:8
Hx10
S:6
tt
Jet multiplicity:
different background
contributions
Higgs Physics at the Tevatron
Wj
Wg
tt
WZ
ZZ
DY
WW
HWWx10
Data
32
No channel too small!
q
W+
W+
UK!
S:1.0 (1A+1B)
S:2.2
q’
l+
n
H
W+
W–
same-sign dilepton
trilepton
S:0.5 ZH
Higgs Physics at the
NNTevatron
Output
l+
n
l– q’
n q
trilepton
S:0.7
WH
NN Output
33
Stability
Assessing NN stability
 rogue variables
– had checked data-simulation agreement in as many regions as possible
Drell Yan-rich
WW-rich
Met
– applicability?
Met
successful training
training epoch
Higgs Physics at the Tevatron
unsuccessful training
training estimator
 convergence
training estimator
 control regions…
training epoch
34
Complementarity
 exploit different sensitivities of matrix element / neural net
– ME is leading order
- remove variables that use jet information from neural net
for comparison
 verify matrix element method: cycle signal
Redefine discriminant for WW hypothesis:
R’ =
PWW
PWW + SkbiPbi
i
Higgs Physics at the Tevatron
35
H→WW
D0: Expect 35 Signal Events
CDF: Expect 32 Signal Events
Bkg uncertainty does not
wash out signal
Higgs Physics at the Tevatron
36
High mass Higgs combination
UK!
PRL 104 061804 (2010)
PRL 104 061803 (2010)
Winter
SM Higgs excluded:
Exp 0.87xSM
Obs 0.93xSM
Higgs061802
Physics at(2010)
the
PRL 104
Tevatron
163<mH<166 GeV
37
New CDF HWW
UK!
Wj
Wg
tt
WZ
ZZ
DY
WW
HWWx10
Data
4.8→5.9 fb-1: 11% better from statistics
Optimized e- selection
Added Trileptons, Low Mll , VH, VBF for 0 Jet
Added t modes
All together: 17% better!
Results at mH = 165GeV: 95%CL Limits/SM
Exp events Exp.
Obs.
Publication 32
1.20 1.29
New
1.00 1.08
Higgs Physics at the Tevatron
39
38
Individual combinations
Higgs Physics at the Tevatron
UK!
39
ICHEP Tevatron combination
UK!
SM Higgs excluded: 158<mH<175 GeV
Higgs Physics at the Tevatron
40
Tevatron projection
Integrated luminosity (pb–1)
End : Sep 2011?
On tape: ~ 8 fb-1 per experiment
Results shown today : 3-6 fb-1
2002
Higgs Physics at the Tevatron
yesterday
41
Higgs Physics at the Tevatron
42
146-183
136-190
Higgs Physics at the Tevatron
43
LHC projections
Higgs Physics at the Tevatron
44
Low mass projections
equivalent luminosity gain
charm discrimination
30%
improved b-tagging
20%
improved dijet mass resolution 15%
extra final states
5–10%
improved lepton id
5–10%
total : 1.4x in the limit
(~ 2x in effective luminosity)
Higgs Physics at the Tevatron
45
Higgs Physics at the Tevatron
46
mH (GeV)
Implications
Altarelli PLB 337 141
mt=174 GeV
log(Λ/GeV) Planck scale
1019 GeV
Higgs Physics at the Tevatron
47
4th generation
g
g
t,X
H
Low-mass scenario:
masses close to
experimental limits
ml4=100GeV
mn4=80GeV
mu4=256GeV
md4=128GeV
Higgs Physics at the Tevatron
48
Cross section x Br 95%CL (pb)
MSSM H→tt
UK!
tan 
Key issue: understanding t ID efficiency
Large calibration samples:
W for ID optimization
Z for efficiency confirmation
Benchmark scenario
No Evidence for SUSY Higgs
tt sensitive at high tan
in future will include
bbtt, bbbb
UK!
Higgs Physics at the Tevatron
49
Ones to watch?
s b jets.Br is factor
2 too large
Higgs Physics at the Tevatron
50
Outlook
♦ Strong physics programs with UK people at the centre.
♦ Many of the primary Tevatron goals have been achieved
– and now we are focused on Higgs physics
♦ Themes: complex analyses, eg taus; combinations
♦ Higgs within reach!
– remarkable times.
Higgs Physics at the Tevatron
51
Higgs Physics at the Tevatron
52
Low mass projections
equivalent luminosity gain
charm discrimination
30%
improved b-tagging
20%
improved dijet mass resolution 15%
extra final states
5–10%
improved lepton id
5–10%
Probability of 3s evidence
Probability of 95% CL Exclusion
total : 1.4x in the limit (~ 2x in effective luminosity)
Higgs Physics at the Tevatron
53
And so many other results:
♦ SUSY limits
♦ limits on new heavy bosons
♦ observations of new baryons
♦ new jet measurements
♦ measurements of as
♦ diffraction
…
Higgs Physics at the Tevatron
54
GW
change to mW
UK!
mW:
CDF: mW = 80413 ± 48 MeV/c2
D0: mW = 80402 ± 43 MeV/c2
Tev: mW = 80420 ± 31 MeV/c2 (includes Run 1)
LEP: mW = 80376 ± 33 MeV/c2
Heading to CDF 25MeV/c2 measurement
Tev error improves from 62 to 49 MeV
Higgs Physics at the Tevatron
55
HWW
Isolation
mll > 16
Dilepton sample
composition
l+
W+
H0
90%
W–
q’
q
q
W+
W–
q’
H0
n
W+
W–
B
Z and top
suppression
l–
n
Z
Z
tt
WW
H
tt
WW
H
signal
separation
l+
n
l–
n
10%
WW
H
WW H
ee, e,  ; ET
Higgs Physics at the Tevatron
56/54
PDFs
p
g
H
p
g
spp→H = sgg→H fg/p(x1,Q=MH) fg/p(x2,Q=MH) + …
Tevatron
y=
2
0
2
LHC
Higgs Physics at the Tevatron
57
Low mass Higgs searches
Higgs Physics at the Tevatron
58
Matrix element method
 Use LO matrix element (MCFM) to compute event probability
xobs:
LO |M|2 :
px
px
py
py
lep1
pz
pz
Ex , E y
lep2
(with true values y)
HWWlnln
WWlnln
ZZllnn
W+partonln+jet
Wgln+g
parton lepton fake rate
g conversion rate
ET model
lepton energy resn
 Compute likelihood ratio discriminator
R=
Ps
Ps + SkbiPbi
kb is relative fraction of expected background contrib.
Ps computed for each mH
i
 Fit templates (separately for high S/B and low S/B dilepton types)
Higgs Physics at the Tevatron
59/54
Neural network method
 Various versions. Current:
 Apply preselection (eg ET to remove Drell-Yan)
 Train on {all backgrounds / WW} against Higgs
mH=110,120…160…200 { possibly separate ee,e, }
score
var1
ET
SET
mll
Elep1
Elep2
ETsig
ETjet1
DRleptons
Dleptons
D ET lep or jet
ETjet2
Njets
Most recent CDF
“combined ME/NN”
analysis also uses
ME LRs as NN input
variables
NN
var2
var n
x10
0
1
Background Higgs
 Pass signal/all backgrounds through net
 Form templates
Data
HWW
WW
DY
Wg
WZ
ZZ
tt
fakes
NN
0
1
 Pass templates and data to fitter
Higgs Physics at the Tevatron
60/54
b-tagging
Secondary vertex-finding algorithm
Attempt to fit tracks to decay vertex
Jet probability
Compares track impact parameters to
measured resolution functions
Neural network filters
ntracks in secondary vertex
pT fraction carried by those tracks
goodness of vertex fit
vertex mass
transverse decay length & significance
…
Higgs Physics at the Tevatron
61/54
Integrated luminosity (fb–1)
Tevatron projection
End : Sep 2011?
On tape: ~ 6 fb-1 per experiment
Results shown today : 3-5 fb-1
Higgs Physics at the Tevatron
62