Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling...

25
Higgs boson self-interactions in SM and BSM (ATLAS+CMS) Roberto Salerno LLR - Ecole Polytechnique - INP23/CNRS

Transcript of Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling...

Page 1: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Higgs boson self-interactions in SM and BSM (ATLAS+CMS)

Roberto Salerno LLR - Ecole Polytechnique - INP23/CNRS

Page 2: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015

Introduction

2

Double Higgs boson (hh) production is the principal way to to study the Higgs boson self-interaction extracting the trilinear coupling (λHHH) → this is generally considered as an analysis for the HL-LHC

*

λg�

g*

gmin

1

04π

λ = √gmin g*─

λ = gmin

FIG. 1: Cartoon of the region in the plane (g⇤,�/g⇤), defined by Eqs. (13),(14), that can be probedby an analysis including only dimension-6 operators (in white). No sensible e↵ective field theorydescription is possible in the gray area (� < gmin), while exploration of the light blue region(gmin < � <

pg⇤gmin) requires including the dimension-8 operators.

g

g h

h

t

g

g h

h

th

g

g h

h

t

g

g h

h

h

g

g h

h

FIG. 2: Feyman diagrams contributing to double Higgs production via gluon fusion (an additionalcontribution comes from the crossing of the box diagram). The last diagram on the first linecontains the tthh coupling, while those in the second line involve contact interactions between theHiggs and the gluons denoted with a cross.

derivative terms (which correspond to dimension-8 operators in the limit of linearly-realized

EW symmetry). The e↵ect of the neglected derivative operators will be then studied by

analyzing their impact on angular di↵erential distributions and shown to be small in our

case due to the limited sensitivity on the high mhh region.

The Feynman diagrams that contribute to the gg ! hh process are shown in Fig. 2. Each

diagram is characterized by a di↵erent scaling at large energiesps = mhh � mt, mh. We

10

*

λg�

g*

gmin

1

04π

λ = √gmin g*─

λ = gmin

FIG. 1: Cartoon of the region in the plane (g⇤,�/g⇤), defined by Eqs. (13),(14), that can be probedby an analysis including only dimension-6 operators (in white). No sensible e↵ective field theorydescription is possible in the gray area (� < gmin), while exploration of the light blue region(gmin < � <

pg⇤gmin) requires including the dimension-8 operators.

g

g h

h

t

g

g h

h

th

g

g h

h

t

g

g h

h

h

g

g h

h

FIG. 2: Feyman diagrams contributing to double Higgs production via gluon fusion (an additionalcontribution comes from the crossing of the box diagram). The last diagram on the first linecontains the tthh coupling, while those in the second line involve contact interactions between theHiggs and the gluons denoted with a cross.

derivative terms (which correspond to dimension-8 operators in the limit of linearly-realized

EW symmetry). The e↵ect of the neglected derivative operators will be then studied by

analyzing their impact on angular di↵erential distributions and shown to be small in our

case due to the limited sensitivity on the high mhh region.

The Feynman diagrams that contribute to the gg ! hh process are shown in Fig. 2. Each

diagram is characterized by a di↵erent scaling at large energiesps = mhh � mt, mh. We

10

λHHH

Marco Zaro, 20-11-2014 LHCPhenoNet

λHHH dependence

5

(a) gg double-Higgs fusion: gg → HH

H

H

H

g

g

Q

H

Hg

g

Q

(b) WW/ZZ double-Higgs fusion: qq′ → HHqq′

q

q′

q

q′

V ∗

V ∗

HH

(c) Double Higgs-strahlung: qq′ → ZHH/WHH

q

q′ V ∗

V

H

H

g

g

t

tHH

q

qg

(d) Associated production with top-quarks: qq/gg → ttHH

Figure 1: Some generic Feynman diagrams contributing to Higgs pair production at hadroncolliders.

where

t± = −s

2

!

1− 2M2

H

s∓"

1−4M2

H

s

#

, (5)

with s and t denoting the partonic Mandelstam variables. The triangular and box formfactors F△, F! and G! approach constant values in the infinite top quark mass limit,

F△ →2

3, F! → −

2

3, G! → 0 . (6)

The expressions with the complete mass dependence are rather lengthy and can be foundin Ref. [11] as well as the NLO QCD corrections in the LET approximation in Ref. [18].

The full LO expressions for F△, F! and G! are used wherever they appear in theNLO corrections in order to improve the perturbative results, similar to what has beendone in the single Higgs production case where using the exact LO expression reduces thedisagreement between the full NLO result and the LET result [7, 19].

For the numerical evaluation we have used the publicly available code HPAIR [44] inwhich the known NLO corrections are implemented. As a central scale for this process

6

(a) gg double-Higgs fusion: gg → HH

H

H

H

g

g

Q

H

Hg

g

Q

(b) WW/ZZ double-Higgs fusion: qq′ → HHqq′

q

q′

q

q′

V ∗

V ∗

HH

(c) Double Higgs-strahlung: qq′ → ZHH/WHH

q

q′ V ∗

V

H

H

g

g

t

tHH

q

qg

(d) Associated production with top-quarks: qq/gg → ttHH

Figure 1: Some generic Feynman diagrams contributing to Higgs pair production at hadroncolliders.

where

t± = −s

2

!

1− 2M2

H

s∓"

1−4M2

H

s

#

, (5)

with s and t denoting the partonic Mandelstam variables. The triangular and box formfactors F△, F! and G! approach constant values in the infinite top quark mass limit,

F△ →2

3, F! → −

2

3, G! → 0 . (6)

The expressions with the complete mass dependence are rather lengthy and can be foundin Ref. [11] as well as the NLO QCD corrections in the LET approximation in Ref. [18].

The full LO expressions for F△, F! and G! are used wherever they appear in theNLO corrections in order to improve the perturbative results, similar to what has beendone in the single Higgs production case where using the exact LO expression reduces thedisagreement between the full NLO result and the LET result [7, 19].

For the numerical evaluation we have used the publicly available code HPAIR [44] inwhich the known NLO corrections are implemented. As a central scale for this process

6

(a) gg double-Higgs fusion: gg → HH

H

H

H

g

g

Q

H

Hg

g

Q

(b) WW/ZZ double-Higgs fusion: qq′ → HHqq′

q

q′

q

q′

V ∗

V ∗

HH

(c) Double Higgs-strahlung: qq′ → ZHH/WHH

q

q′ V ∗

V

H

H

g

g

t

tHH

q

qg

(d) Associated production with top-quarks: qq/gg → ttHH

Figure 1: Some generic Feynman diagrams contributing to Higgs pair production at hadroncolliders.

where

t± = −s

2

!

1− 2M2

H

s∓"

1−4M2

H

s

#

, (5)

with s and t denoting the partonic Mandelstam variables. The triangular and box formfactors F△, F! and G! approach constant values in the infinite top quark mass limit,

F△ →2

3, F! → −

2

3, G! → 0 . (6)

The expressions with the complete mass dependence are rather lengthy and can be foundin Ref. [11] as well as the NLO QCD corrections in the LET approximation in Ref. [18].

The full LO expressions for F△, F! and G! are used wherever they appear in theNLO corrections in order to improve the perturbative results, similar to what has beendone in the single Higgs production case where using the exact LO expression reduces thedisagreement between the full NLO result and the LET result [7, 19].

For the numerical evaluation we have used the publicly available code HPAIR [44] inwhich the known NLO corrections are implemented. As a central scale for this process

6

(a) gg double-Higgs fusion: gg → HH

H

H

H

g

g

Q

H

Hg

g

Q

(b) WW/ZZ double-Higgs fusion: qq′ → HHqq′

q

q′

q

q′

V ∗

V ∗

HH

(c) Double Higgs-strahlung: qq′ → ZHH/WHH

q

q′ V ∗

V

H

H

g

g

t

tHH

q

qg

(d) Associated production with top-quarks: qq/gg → ttHH

Figure 1: Some generic Feynman diagrams contributing to Higgs pair production at hadroncolliders.

where

t± = −s

2

!

1− 2M2

H

s∓"

1−4M2

H

s

#

, (5)

with s and t denoting the partonic Mandelstam variables. The triangular and box formfactors F△, F! and G! approach constant values in the infinite top quark mass limit,

F△ →2

3, F! → −

2

3, G! → 0 . (6)

The expressions with the complete mass dependence are rather lengthy and can be foundin Ref. [11] as well as the NLO QCD corrections in the LET approximation in Ref. [18].

The full LO expressions for F△, F! and G! are used wherever they appear in theNLO corrections in order to improve the perturbative results, similar to what has beendone in the single Higgs production case where using the exact LO expression reduces thedisagreement between the full NLO result and the LET result [7, 19].

For the numerical evaluation we have used the publicly available code HPAIR [44] inwhich the known NLO corrections are implemented. As a central scale for this process

6

10-1

100

101

102

-4 -3 -2 -1 0 1 2 3 4

m(N

)LO

[fb]

h/hSM

ppAHH (EFT loop-improved)ppAHHjj (VBF)

ppAttHH

ppAWHH

ppAZHH ppAtjHH

HH production at 14 TeV LHC at (N)LO in QCDMH=125 GeV, MSTW2008 (N)LO pdf (68%cl)

MadGraph5_

aMC@NLO

SM

arXiv:1401.7340 arXiv:1408.6542

-4 -3 -2 -1 0 1 2 3 4λHHH/λHHHSM

σ (N

)LO

[fb]

10-1

1

10

102

√s σ (fb)7 TeV 6.858 TeV 9.96

13 TeV 34.314 TeV 40.7

Preliminary recommendations from the hh group of LHXSWG

σSM(gg→hh) mh = 125 GeV

gluon fusion channel

Page 3: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015

Introduction

3

Double Higgs boson (hh) production is the principal way to to study the Higgs boson self-interaction extracting the trilinear coupling (λHHH) → this is generally considered as an analysis for the HL-LHC

Even if in LHC Run1 we did not have any sensitivity to “measure” Standard Model λHHH an easy pattern to look for hh are : → resonant production from decay of new exotic particles

→ non-resonant production from SM or from new diagrams increasing the production cross section

• non SM Yukawa couplings • ttHH interactions • dimension-6 gluon Higgs operators • light coloured scalars • …

*

λg�

g*

gmin

1

04π

λ = √gmin g*─

λ = gmin

FIG. 1: Cartoon of the region in the plane (g⇤,�/g⇤), defined by Eqs. (13),(14), that can be probedby an analysis including only dimension-6 operators (in white). No sensible e↵ective field theorydescription is possible in the gray area (� < gmin), while exploration of the light blue region(gmin < � <

pg⇤gmin) requires including the dimension-8 operators.

g

g h

h

t

g

g h

h

th

g

g h

h

t

g

g h

h

h

g

g h

h

FIG. 2: Feyman diagrams contributing to double Higgs production via gluon fusion (an additionalcontribution comes from the crossing of the box diagram). The last diagram on the first linecontains the tthh coupling, while those in the second line involve contact interactions between theHiggs and the gluons denoted with a cross.

derivative terms (which correspond to dimension-8 operators in the limit of linearly-realized

EW symmetry). The e↵ect of the neglected derivative operators will be then studied by

analyzing their impact on angular di↵erential distributions and shown to be small in our

case due to the limited sensitivity on the high mhh region.

The Feynman diagrams that contribute to the gg ! hh process are shown in Fig. 2. Each

diagram is characterized by a di↵erent scaling at large energiesps = mhh � mt, mh. We

10

*

λg�

g*

gmin

1

04π

λ = √gmin g*─

λ = gmin

FIG. 1: Cartoon of the region in the plane (g⇤,�/g⇤), defined by Eqs. (13),(14), that can be probedby an analysis including only dimension-6 operators (in white). No sensible e↵ective field theorydescription is possible in the gray area (� < gmin), while exploration of the light blue region(gmin < � <

pg⇤gmin) requires including the dimension-8 operators.

g

g h

h

t

g

g h

h

th

g

g h

h

t

g

g h

h

h

g

g h

h

FIG. 2: Feyman diagrams contributing to double Higgs production via gluon fusion (an additionalcontribution comes from the crossing of the box diagram). The last diagram on the first linecontains the tthh coupling, while those in the second line involve contact interactions between theHiggs and the gluons denoted with a cross.

derivative terms (which correspond to dimension-8 operators in the limit of linearly-realized

EW symmetry). The e↵ect of the neglected derivative operators will be then studied by

analyzing their impact on angular di↵erential distributions and shown to be small in our

case due to the limited sensitivity on the high mhh region.

The Feynman diagrams that contribute to the gg ! hh process are shown in Fig. 2. Each

diagram is characterized by a di↵erent scaling at large energiesps = mhh � mt, mh. We

10

λHHHgluon fusion channel

Page 4: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015

Which final states?

4

Branching ratios and production mechanisms are decoupled effects Double Higgs boson production has a phenomenologically rich set of final states

No

trev

iew

ed

,fo

rin

tern

al

circu

la

tio

no

nly

Draft version 1.0

ATLAS NOTEFebruary 26, 2013

Study of the spin of the Higgs-like particle in the H ! WW(⇤) ! e⌫µ⌫1

channel with 20.7 fb�1 of⇧

s = 8 TeV data collected with the ATLAS2

detector3

The ATLAS Collaboration4

Abstract5

Recently, the ATLAS collaboration reported the observation of a new neutral particle6

in the search for the Standard Model Higgs boson. The measured production rate of the7

new particle is consistent with the Standard Model Higgs boson with a mass of about 1258

GeV, but its other physics properties are unknown. Presently, the only constraint on the9

spin of this particle stems from the observed decay mode to two photons, which disfavours10

a spin-1 hypothesis. This note reports on the compatibility of the observed excess in the11

H ⌅ WW(⇥) ⌅ e⇥µ⇥ search arising from either a spin-0 or a spin-2 particle with positive12

charge-parity. Data collected in 2012 with the ATLAS detector favours a spin-0 signal, and13

results in the exclusion of a spin-2 signal at 95% confidence level if one assumes a qq ⌅ X14

production fraction larger than 25% for a spin-2 particle, and at 91% confidence level if one15

assumes pure gg production.16

c⇤ Copyright 2013 CERN for the benefit of the ATLAS Collaboration.Reproduction of this article or parts of it is allowed as specified in the CC-BY-3.0 license.

- John Alison - Experimental Studies of hh Higgs Coupling 2014

hh Decay

13

-810

-710

-610

-510

-410

-310

-210

-110

1

bb WW gg oo cc ZZ aa aZ µµ

µµ

aZ

aa

ZZ

cc

oo

ggWW

bb 33%

3e-3

7%

25%

Phenomenologically rich set of final states. hh-Br

- John Alison - Higgs Coupling 2014

Larger Br-h decay

Rarer Br-h decay

→SM branching ratios (for mh = 125 GeV) are used as first approximation for all the analyses

Larg

er B

R

Rarer BR

2×BRSM(h→xx)×BRSM(h→yy)

[BRSM(h→xx)]2if xx ≠ yy

if xx = yy

Covered final state

by ATLAS/CMS

Page 5: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

(X)→hh→bbbb

BRSM(hh→bbbb) ~ 33.3%

Phys. Lett. B 749 (2015) 560

CMS-EXO-12-053 Eur. Phys. J. C (2015) 75:412

No

trev

iew

ed

,fo

rin

tern

al

circu

latio

no

nly

Draft version 1.0

ATLAS NOTEFebruary 26, 2013

Study of the spin of the Higgs-like particle in the H ! WW(⇤) ! e⌫µ⌫1

channel with 20.7 fb�1 of⇧

s = 8 TeV data collected with the ATLAS2

detector3

The ATLAS Collaboration4

Abstract5

Recently, the ATLAS collaboration reported the observation of a new neutral particle6

in the search for the Standard Model Higgs boson. The measured production rate of the7

new particle is consistent with the Standard Model Higgs boson with a mass of about 1258

GeV, but its other physics properties are unknown. Presently, the only constraint on the9

spin of this particle stems from the observed decay mode to two photons, which disfavours10

a spin-1 hypothesis. This note reports on the compatibility of the observed excess in the11

H ⌅ WW(⇥) ⌅ e⇥µ⇥ search arising from either a spin-0 or a spin-2 particle with positive12

charge-parity. Data collected in 2012 with the ATLAS detector favours a spin-0 signal, and13

results in the exclusion of a spin-2 signal at 95% confidence level if one assumes a qq ⌅ X14

production fraction larger than 25% for a spin-2 particle, and at 91% confidence level if one15

assumes pure gg production.16

c⇤ Copyright 2013 CERN for the benefit of the ATLAS Collaboration.Reproduction of this article or parts of it is allowed as specified in the CC-BY-3.0 license.

- John Alison - Experimental Studies of hh Higgs Coupling 2014

hh Decay

13

-810

-710

-610

-510

-410

-310

-210

-110

1

bb WW gg oo cc ZZ aa aZ µµ

µµ

aZ

aa

ZZ

cc

oo

ggWW

bb 33%

3e-3

7%

25%

Phenomenologically rich set of final states. hh-Br

- John Alison - Higgs Coupling 2014

Larger Br-h decay

Rarer Br-h decay

Page 6: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015

HH(4b), Event Selections

‣ The trigger requires ‣ 4 anti-kT (0.5) central jets with pT > 30 GeV ‣ 2 with pT > 80 GeV ‣ 2 b-tag

‣ 4 central jets with pT > 40 GeV and CMVA>0.71

‣ Two Higgs candidates

‣ m(bb) ~ mH ± 35 GeV

‣ at least 2 jets with pT > 90 GeV

Low Mass Regime2mH < mX < 450 GeV

‣ ΔR(bb)<1.5

Medium Mass Regime450 GeV < mX < 750 GeV

‣ ΔR(bb)<1.5 and pT(bb) > 300 GeV High Mass RegimemX >700 GeV

‣ Signal Region

(GeV)Xm200 400 600 800 1000

-410

-310

-210

-110

1

10

Trigger and b-tag

TJet pHH candidateSignal Region

(8 TeV)-117.93 fb

CMSPreliminary Simulation

4

(GeV) X m400 500 600 700 800 900

Eve

nts

/ 10

GeV

0

5

10

15

20

25

/ n = 0.962χ p-value = 0.56

Data in SRBackgroundMultijetttSignal

(8 TeV)-117.9 fb

CMS

(GeV) Xm300 350 400 450 500 550 600

Even

ts /

10 G

eV

0

10

20

30

40

50

/ n = 0.802χ p-value = 0.96

Data in SRBackgroundMultijetttSignal

(8 TeV)-117.9 fb

CMS

(GeV)X m600 700 800 900 1000 1100 1200

Eve

nts

/ 50

GeV

0

2

4

6

8

10

12

14

16

18

/ n = 1.492χ p-value = 0.13

Data in SRBackgroundMultijetttSignal

(8 TeV)-117.9 fb

CMS

6

A model-independent search for a narrow-width resonance in 270-1100 GeV range

4 anti-kT ΔR=0.5 b-jets with pT > 40

The kinematic distributions of the decay products vary substantially over the mass range 3 kinematical regions A kinematic fit is performed to improve the mass resolution mX resolution is 4-30 GeV (improves by 20-40%) 450 730

(X)→hh→bbbb : resolved analysis Phys. Lett. B 749 (2015) 560CMS

low-mass medium-mass high-mass

Page 7: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/20157

300 and 1100 GeV are excluded at 95% CL 380 and 830 GeV are excluded at 95% CL

(X)→hh→bbbb : resolved analysis CMS

The exclusion limits for radion production and Kaluza–Klein graviton production

(GeV)X m200 300 400 500 600 700 800 900 1000 1100

) (fb

)bbb

b→

HH

X

→ (p

p σ

lim

it on

S

95%

CL

1

10

210

310

410

(8 TeV)-117.9 fb

CMS Expected limit σ 1 ±Expected σ 2 ±Expected

Observed limitKK-Graviton

kL = 35, no R/H mixing

different acceptance+20-30% signal efficiency

(GeV)X m200 300 400 500 600 700 800 900 1000 1100

) (fb

)bbb

b→

HH

X

→ (p

p σ

lim

it on

S

95%

CL

1

10

210

310

410

(8 TeV)-117.9 fb

CMS

kL = 35, no R/H mixing

Expected limit σ 1 ±Expected σ 2 ±Expected

Observed limit = 1 TeVRΛRadion

The low mass region sensitivity not enough to probe (N)MSSM predictions

Phys. Lett. B 749 (2015) 560

Page 8: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/20158

Above mX > 1TeV significant merging takes place → one massive jet instead of two resolved b jets → jet sub-structure techniques exploited to perform H-tag and jet quality

(X)→hh→bbbb : boosted analysis CMS-PAS-EXO-12-053 CMS

Starting with 2 CA ΔR=0.8 central jets

Resonance Mass (TeV)1 1.5 2 2.5 3

H(b

b)H

(bb)

) (fb

)→

BR

(X

× σ

-110

1

10

210

310

410 (8 TeV)-119.7 fb

CMSPreliminary Observed

Expectedσ 1 ±Expected σ 2 ±Expected

= 1 TeV)RΛRadion ( = 3 TeV)RΛRadion (

Resonance Mass (TeV)1 1.5 2 2.5 3

H(b

b)H

(bb)

) (fb

)→

BR

(X

× σ

-110

1

10

210

310

410 (8 TeV)-119.7 fb

CMSPreliminary Observed

Expectedσ 1 ±Expected σ 2 ±Expected

= 1 TeV)RΛRadion ( = 3 TeV)RΛRadion (

Radion (ΛR = 1TeV)

Radion (ΛR = 3TeV)

Resonance Mass (TeV)1 1.5 2 2.5 3

H(b

b)H

(bb)

) (fb

)→

BR

(X

× σ

-110

1

10

210

310

410 (8 TeV)-119.7 fb

CMSPreliminary Observed

Expectedσ 1 ±Expected σ 2 ±Expected

= 1 TeV)RΛRadion ( = 3 TeV)RΛRadion (

(GeV)jjm1000120014001600180020002200

Even

ts /

GeV

-310

-210

-110

1

10) with HP (HP)2 (Jet1Jet

Data

Bckg Fit

Bckg Uncertainty

= 3 TeVRΛSignal for

1× = 1.5 TeV XM

5× = 2.0 TeV XM

bbb b→ HH →X

(8 TeV)-119.7 fbPreliminaryCMS

(GeV)jjm1000 1200 1400 1600 1800 2000 2200

D

ata

σD

ata

- Bac

kg

-1.5-1.0-0.50.00.51.01.5

21τN-subjetiness ratio 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Even

ts /

0.05

0

500

1000

1500

2000

2500

310×

Data before H tagging

Multijet (MADGRAPH+PYTHIA)

160,000) (MADGRAPH+PYTHIA)× HH (→Radion (1.5TeV)

4,000,000) (MADGRAPH+PYTHIA)× HH (→Radion (2.5TeV)

CA8 pruned R=0.8

CMSPreliminary

(8 TeV)-119.7 fb

High-purity Low-purity

(GeV)jPPruned-jet mass m

0 20 40 60 80 100 120 140 160 180 200

Even

ts /

5 G

eV

0

500

1000

1500

2000

2500

310×

Data before H tagging

Multijet (MADGRAPH+PYTHIA)

160,000) (MADGRAPH+PYTHIA)× HH (→Radion (1.5TeV)

4,000,000) (MADGRAPH+PYTHIA)× HH (→Radion (2.5TeV)

CA8 pruned R=0.8

CMSPreliminary

(8 TeV)-119.7 fb

H-tag jet quality

Page 9: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015

[GeV]KKG*m

600 800 1000 1200 1400

Acce

ptan

ce x

Effi

cien

cy

0

0.05

0.1

0.15

0.24 b-tagged jets2 dijetsMass Dependent Cuts

VetottSignal Region

= 1.0PlMBulk RS, k/ATLAS Simulation

= 8 TeVs

(a) Bulk RS, k/MPl = 1

[GeV]Hm600 800 1000 1200 1400

Acce

ptan

ce x

Effi

cien

cy

0

0.05

0.1

0.15

0.24 b-tagged jets2 dijetsMass Dependent Cuts

VetottSignal Region

= 1 GeVHΓ2HDM, with fixed ATLAS Simulation

= 8 TeVs

(b) pp! H! hh! bbbb with fixed �H = 1 GeV

Figure 2: The selection e�ciency as a function of resonance mass at each stage of the event selection for (a) G⇤KK! hh! bbbb events and (b) H! hh! bbbb events in the resolved analysis.

the spin-2 G⇤KK decay (due to the di↵ering angular distributions of hh), resulting in lower acceptance forthese kinematic requirements at low resonance mass.

The resolved analysis considers a large range of resonance masses, 500 mX 1500 GeV. Due tothe di↵ering kinematics, the optimal selection for low-mass resonances di↵ers from the optimum forhigher masses. To increase the analysis sensitivity, three requirements which vary with the reconstructedresonance mass are used. These selection requirements are optimized simultaneously, by performing athree-dimensional scan of threshold values, using the statistical-only exclusion limit (Sect. 6.2) as theobjective function. There are mass-dependent requirements (shown in Fig. 2 as “MDC”) on the minimumpT of the leading and subleading dijets as well as on the maximum di↵erence in pseudorapidity,

����⌘dijets���,

between them. These requirements are written in terms of the four-jet mass m4j expressed in GeV:

pleadT >

8>>>>><>>>>>:

400 GeV if m4j > 910 GeV,200 GeV if m4j < 600 GeV,0.65m4j � 190 GeV otherwise,

psublT >

8>>>>><>>>>>:

260 GeV if m4j > 990 GeV,150 GeV if m4j < 520 GeV,0.235m4j + 28 GeV otherwise,

����⌘dijets��� <8>><>>:

1 if m4j < 820 GeV,1.55 ⇥ 10�3m4j � 0.27 otherwise.

The di↵erent m4j thresholds shown above are chosen to obtain a continuously varying set of requirements.The requirement on

����⌘dijets��� leads to a lower acceptance for H compared to G⇤KK for mX � 700 GeV

because of the e↵ect of the boson spin on the angular distribution of its decay products.

After selecting two dijets that satisfy the mass-dependent criteria, tt constitutes approximately 10% of thetotal background. This tt background predominantly comprises events where both top quarks decayedhadronically. These hadronic decays lead to three jets for each top quark: one b-jet directly from the top

7

9

[GeV]4jm400 600 800 1000 1200 1400 1600

Even

ts /

20 G

eV

2

4

6

8

10

12

14

16

18ATLAS

= 8 TeVs-1Ldt = 19.5 fb∫

Signal RegionDataMultijettt

Syst+Stat Uncertainty = 1.0PlMG*(700), k/

3× = 1.0, PlMG*(1000), k/

[GeV]4jm400 600 800 1000 1200 1400 1600D

ata

/ Bkg

d

012345

(X)→hh→bbbb : resolved analysis ATLAS

4 anti-kT ΔR=0.4 b-jets with pT > 40 Di-jet system (pT

Sublead> 150 && pTLead > 200)

Mass dependent cuts in 3 regions tt Veto Signal region with mass depended cuts

Background Multijet (~95%) from sidebands tt yields from data shape from MC

Analysis used to set a limit on the non-resonant SM hh production

Eur. Phys. J. C (2015) 75:412

Page 10: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015

[GeV]leadJm

0 50 100 150 200 250 300

[G

eV

]su

bl

Jm

50

100

150

200

250

300 2E

ven

ts /

16

Ge

V

0

10

20

30

40

50

60ATLAS-1

Ldt = 19.5 fb∫ = 8 TeV s

Figure 8: The leading–subleading large-R jet mass distribution for the 2-tag and 3-tag data sample in the boostedanalysis. The signal region is the area surrounded by the inner black contour line, centred on mlead

J = 124 GeV andmsubl

J = 115 GeV. The control region is the area inside the outer black contour line, excluding the signal region. Thesideband region is the area outside the outer contour line.

The estimated background yield in the 4-tag sample, N4�tagbkg , is computed according to

N4�tagbkg = µQCD N2+3�tag

QCD + ↵tt N4�tagtt + N4�tag

Z , (4)

where N2+3�tagQCD is the number of multijet events in the 2+3-tag data sample, N4�tag

tt and N4�tagZ are the

numbers of events in the 4-tag tt and Z+jets MC samples. The parameter µQCD corresponds to the ratio ofmultijet event yields in the 4-tag and 2+3-tag data samples, as defined in Eq. (2), except for including both2- and 3-tag events in the denominator. Finally, the parameter ↵tt is a scale factor designed to adjust the ttevent yield from the MC simulation. Both µQCD and ↵tt are extracted from a binned likelihood fit to theleading large-R jet mass distribution obtained in the sideband region of the 4-tag data sample, as depictedin Fig. 9. Due to the large minimum pT requirement for the leading large-R jet, much of the tt contributionis concentrated at high mass close to the top-quark mass. In this fit, the multijet distribution is extractedfrom the 2+3-tag data sample, after subtraction of the tt and Z+jets contributions predicted by the MCsimulation. The tt and Z+jets distributions in the sideband region of the 4-tag data sample are taken fromthe MC simulation, but the Z+jets contribution is very small and its distribution is added to the multijetdistribution for the fit. The resulting fit values are µQCD = 0.0071 ± 0.0007 and ↵tt = 1.44 ± 0.50 with acorrelation coe�cient of �0.67 between these two parameters.

Figure 10a shows the dijet mass distribution for the 4-tag data sample in the sideband region with thebackground estimated using the above method. This figure indicates that the 2+3-tag sample provides avalid description of the background kinematics in the 4-tag sample. The modelling of the background

18

10

[GeV]2Jm600 800 1000 1200 1400 1600 1800 2000D

ata

/ Bkg

d

012345

[GeV]2Jm600 800 1000 1200 1400 1600 1800 2000

Even

ts /

50 G

eV

2

4

6

8

10

12Signal Region

DataMultijettt

Syst+Stat Uncertainty 2× = 1.0, PlMG*(1000), k/ 15× = 1.0, PlMG*(1500), k/

ATLAS = 8 TeVs

-1Ldt = 19.5 fb∫

2 anti-kT ΔR=1.0 jets with pT > 250 GeV Trimming to remove pile-up effects Track jets b-tag (ΔR=0.3) Use jet mass to test Higgs mass compatibility

(X)→hh→bbbb : boosted analysis Eur. Phys. J. C (2015) 75:412 ATLAS

Background Multijet (~90%) from sidebands tt yields from data shape from MC

Signal Region

Page 11: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015

[GeV]KKG*m

600 800 1000 1200 1400 1600 1800 2000

) [fb

]bbb

b→

hh

→ KK

G*

→(p

p σ

1

10

210

= 1.0PlMBulk RS, k/Observed Limit (95% CL)Expected Limit (95% CL)

σ1±Expected σ2±Expected

ATLAS-1 Ldt = 19.5 fb∫ = 8 TeV s

(a) Bulk RS, k/MPl = 1

[GeV]KKG*m

600 800 1000 1200 1400 1600 1800 2000

) [fb

]bbb

b→

hh

→ KK

G*

→(p

p σ

1

10

210

= 2.0PlMBulk RS, k/Observed Limit (95% CL)Expected Limit (95% CL)

σ1±Expected σ2±Expected

ATLAS-1 Ldt = 19.5 fb∫ = 8 TeV s

(b) Bulk RS, k/MPl = 2

[GeV]Hm600 800 1000 1200 1400

) [fb

]bbb

b→

hh

→ H

(pp

σ

1

10

210

Observed Limit (95% CL)Expected Limit (95% CL)

σ1±Expected σ2±Expected

ATLAS-1 Ldt = 19.5 fb∫ = 8 TeV s

= 1 GeVHΓ

(c) pp ! H! hh! bbbb with fixed �H =

1 GeV

Figure 14: The combined expected and observed limit for pp! G⇤KK! hh! bbbb in the bulk RS model with (a)k/MPl = 1 and (b) k/MPl = 2, as well as (c) pp! H! hh! bbbb with fixed �H = 1 GeV. The red curves show thepredicted cross-sections as a function of resonance mass for the models considered.

Table 9: Range of KK graviton masses excluded at 95% confidence level for k/MPl = 1.0, 1.5 and 2.0.

k/MPl 95% CL Excluded G⇤KK Mass Range [GeV]

1.0 500 � 7201.5 500 � 800 and 870 � 9102.0 500 � 990

where there is little expected background and either the resolved or boosted analysis provides good signalacceptance. The excluded mass ranges for the bulk RS KK graviton are shown in Table 9.

The excluded mass range for the 2HDM is parameter dependent, principally because the productioncross-section varies, but also because the exclusion limit depends on the parameter-dependent H bosonwidth, �H . The theoretical cross-section used to determine the 95% CL excluded regions is the sum of thecross-sections of gluon-fusion production, vector-boson-fusion production and b-associated production.

The e↵ects of �H are accounted for by creating mH distributions with a range of widths, 0 < �H/mH 0.5,for each mH considered. These distributions are based on parameterizations which include resolution andacceptance e↵ects combined with a Breit–Wigner line-shape. A grid of limits are calculated with eachof these mass distributions. Then, for each point in mH , cos (� � ↵), and tan � space, the cross-sectionlimit is determined by interpolating between the appropriate limits, based on the �H given by the modelfor that point. For the widest signals considered, the exclusion limits worsen by up to a factor of three.The exclusion regions determined through this process are shown as a function of cos (� � ↵) and tan � formH = 500 GeV in Figs. 15 and 16, and as a function of mH and tan � for cos (� � ↵) = �0.2 in Figs. 17and 18. The validity of the process has been tested using the widest available signals, gravitons in the bulkRS model with k/MPl = 2. Phase-space regions with �H greater than these graviton widths are consideredunvalidated and are shown in the figures as grey areas.

25

11

(X)→hh→bbbb : results

Excluded regions are presented in four 2HDMs (Type-I, Type-II, Lepton-specific, Flipped)

The exclusion limit for pp→H→hh→bbbb with fixed ΓH=1GeV

A simple combination of the separate limits from the resolved and boosted (performance crossing around 1.1 TeV)

boostedresolved

ATLAS Eur. Phys. J. C (2015) 75:412

Page 12: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

(X)→hh→bb𝝉𝝉

BRSM(hh→bb𝝉𝝉) ~ 7.2%

arXiv:1510.01181 arXiv:1509.04670

No

trev

iew

ed

,fo

rin

tern

al

circu

latio

no

nly

Draft version 1.0

ATLAS NOTEFebruary 26, 2013

Study of the spin of the Higgs-like particle in the H ! WW(⇤) ! e⌫µ⌫1

channel with 20.7 fb�1 of⇧

s = 8 TeV data collected with the ATLAS2

detector3

The ATLAS Collaboration4

Abstract5

Recently, the ATLAS collaboration reported the observation of a new neutral particle6

in the search for the Standard Model Higgs boson. The measured production rate of the7

new particle is consistent with the Standard Model Higgs boson with a mass of about 1258

GeV, but its other physics properties are unknown. Presently, the only constraint on the9

spin of this particle stems from the observed decay mode to two photons, which disfavours10

a spin-1 hypothesis. This note reports on the compatibility of the observed excess in the11

H ⌅ WW(⇥) ⌅ e⇥µ⇥ search arising from either a spin-0 or a spin-2 particle with positive12

charge-parity. Data collected in 2012 with the ATLAS detector favours a spin-0 signal, and13

results in the exclusion of a spin-2 signal at 95% confidence level if one assumes a qq ⌅ X14

production fraction larger than 25% for a spin-2 particle, and at 91% confidence level if one15

assumes pure gg production.16

c⇤ Copyright 2013 CERN for the benefit of the ATLAS Collaboration.Reproduction of this article or parts of it is allowed as specified in the CC-BY-3.0 license.

- John Alison - Experimental Studies of hh Higgs Coupling 2014

hh Decay

13

-810

-710

-610

-510

-410

-310

-210

-110

1

bb WW gg oo cc ZZ aa aZ µµ

µµ

aZ

aa

ZZ

cc

oo

ggWW

bb 33%

3e-3

7%

25%

Phenomenologically rich set of final states. hh-Br

- John Alison - Higgs Coupling 2014

Larger Br-h decay

Rarer Br-h decay

Page 13: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/201513

Final states: 𝝉h𝝉h e𝝉h μ𝝉h

divided in categories based on number of b-jets (0 or 1 or 2)

(X)→hh→bb𝝉𝝉CMS

11

(GeV)kinfitHM

200 300 400 500 600 700

(1/G

eV)

kinf

itH

dN/d

M

0

2

4

6

8

10

12 CMS (8 TeV)-119.7 fb

, 2jet-0taghτe

scenarioβMSSM low tan= 2β= 300 GeV, tanHm

ττbb→hh→H×10Observed

ττ→Z ee→Z

W+jetsQCDtt

Bkg. uncertainty

(GeV)kinfitHM

200 300 400 500 600 700

(1/G

eV)

kinf

itH

dN/d

M

0.0

0.5

1.0

1.5

2.0

2.5

3.0

CMS (8 TeV)-119.7 fb

, 2jet-1taghτe

scenarioβMSSM low tan= 2β= 300 GeV, tanHm

ττbb→hh→H×10Observed

ττ→Z ee→Z

W+jetsQCDtt

Bkg. uncertainty

(GeV)kinfitHM

200 300 400 500 600 700

(1/G

eV)

kinf

itH

dN/d

M

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9 CMS (8 TeV)-119.7 fb

, 2jet-2taghτe

scenarioβMSSM low tan= 2β= 300 GeV, tanHm

ττbb→hh→H×10Observed

ττ→Z ee→Z

W+jetsQCDtt

Bkg. uncertainty

Figure 4: Distributions of the reconstructed four-body mass with the kinematic fit after apply-ing mass selections on mtt and mbb in the eth channel. The plots are shown for events in the2jet–0tag (top left), 2jet–1tag (top right), and 2jet–2tag (bottom) categories. The expected signalscaled by a factor 10 is shown superimposed as an open dashed histogram for tan b = 2 andmH = 300 GeV in the low tan b scenario of the MSSM. Expected background contributions areshown for the values of nuisance parameters (systematic uncertainties) obtained after fittingthe signal plus background hypothesis to the data.

10 7 Results and interpretation

(GeV)kinfitHM

200 300 400 500 600 700

(1/G

eV)

kinf

itH

dN/d

M

0

2

4

6

8

10

12

14

16

18

20

CMS (8 TeV)-119.7 fb

, 2jet-0taghτµ

scenarioβMSSM low tan= 2β= 300 GeV, tanHm

ττbb→hh→H×10Observed

ττ→ZElectroweakQCDtt

Bkg. uncertainty

(GeV)kinfitHM

200 300 400 500 600 700

(1/G

eV)

kinf

itH

dN/d

M

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

CMS (8 TeV)-119.7 fb

, 2jet-1taghτµ

scenarioβMSSM low tan= 2β= 300 GeV, tanHm

ττbb→hh→H×10Observed

ττ→ZElectroweakQCDtt

Bkg. uncertainty

(GeV)kinfitHM

200 300 400 500 600 700

(1/G

eV)

kinf

itH

dN/d

M

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8 CMS (8 TeV)-119.7 fb

, 2jet-2taghτµ

scenarioβMSSM low tan= 2β= 300 GeV, tanHm

ττbb→hh→H×10Observed

ττ→ZElectroweakQCDtt

Bkg. uncertainty

Figure 3: Distributions of the reconstructed four-body mass with the kinematic fit after apply-ing mass selections on mtt and mbb in the µth channel. The plots are shown for events in the2jet–0tag (top left), 2jet–1tag (top right), and 2jet–2tag (bottom) categories. The expected signalscaled by a factor 10 is shown superimposed as an open dashed histogram for tan b = 2 andmH = 300 GeV in the low tan b scenario of the MSSM. Expected background contributions areshown for the values of nuisance parameters (systematic uncertainties) obtained after fittingthe signal plus background hypothesis to the data.

[GeV]Hm260 280 300 320 340

bb) [

pb]

ττ→

hh→

BR

(H×

H)

→(g

95%

CL

limit

on

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4 CMSUnpublished

(8 TeV)-119.7 fbbbττ→hh→H

Expected limit

hτe

hτµ

hτhτ

Combined

12 7 Results and interpretation

(GeV)kinfitHM

200 300 400 500 600 700

(1/G

eV)

kinf

itH

dN/d

M

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

CMS (8 TeV)-119.7 fb

, 2jet-0taghτhτ

scenarioβMSSM low tan= 2β= 300 GeV, tanHm

ττbb→hh→ H×10 Observed

ττ→ZElectroweakQCDtt

Bkg. uncertainty

(GeV)kinfitHM

200 300 400 500 600 700

(1/G

eV)

kinf

itH

dN/d

M

0.0

0.1

0.2

0.3

0.4

0.5

0.6 CMS (8 TeV)-119.7 fb

, 2jet-1taghτhτ

scenarioβMSSM low tan= 2β= 300 GeV, tanHm

ττbb→hh→ H×10 Observed

ττ→ZElectroweakQCDtt

Bkg. uncertainty

(GeV)kinfitHM

200 300 400 500 600 700

(1/G

eV)

kinf

itH

dN/d

M

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35 CMS (8 TeV)-119.7 fb

, 2jet-2taghτhτ

scenarioβMSSM low tan= 2β= 300 GeV, tanHm

ττbb→hh→ H×10 Observed

ττ→ZElectroweakQCDtt

Bkg. uncertainty

Figure 5: Distributions of the reconstructed four-body mass with the kinematic fit after apply-ing mass selections on mtt and mbb in the thth channel. The plots are shown for events in the2jet–0tag (top left), 2jet–1tag (top right), and 2jet–2tag (bottom) categories. The expected signalscaled by a factor 10 is shown superimposed as an open dashed histogram for tan b = 2 andmH = 300 GeV in the low tan b scenario of the MSSM. Expected background contributions areshown for the values of nuisance parameters (systematic uncertainties) obtained after fittingthe signal plus background hypothesis to the data.

𝝉h𝝉h e𝝉h

μ𝝉h

Selection largely following the SM H→𝝉𝝉 analysis

Kinematical fit (MHkinfit) for Mhh signal-to-background ratio is greatly improved

Comparison of the expected limits separated by final states

Background (tt,QCD,Z→𝝉𝝉,…) shapes/yields mainly from data

arXiv:1510.01181

Page 14: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/201514

15

(GeV)Hm260 280 300 320 340

) (pb

)ττ

bb→

hh→

B(H

×(g

gH)

σ95

% C

L lim

it on

0.2

0.4

0.6

0.8

1.01.2

1.4

1.6

1.8

2.0ObservedExpected

Expectedσ 1± Expectedσ 2±

CMS (8 TeV)-119.7 fb

hτµ

ττbb→hh→H

(GeV)Hm260 280 300 320 340

) (pb

)ττ

bb→

hh→

B(H

×(g

gH)

σ95

% C

L lim

it on

0.2

0.4

0.6

0.8

1.01.2

1.4

1.6

1.8

2.0ObservedExpected

Expectedσ 1± Expectedσ 2±

CMS (8 TeV)-119.7 fb

hτe

ττbb→hh→H

(GeV)Hm260 280 300 320 340

) (pb

)ττ

bb→

hh→

B(H

×(g

gH)

σ95

% C

L lim

it on

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0ObservedExpected

Expectedσ 1± Expectedσ 2±

CMS (8 TeV)-119.7 fb

hτhτ

ττbb→hh→H

(GeV)Hm260 280 300 320 340

) (pb

)ττ

bb→

hh→

B(H

×(g

gH)

σ95

% C

L lim

it on

0.2

0.4

0.6

0.8

1.0

1.2

1.4ObservedExpected

Expectedσ 1± Expectedσ 2±

CMS (8 TeV)-119.7 fb

hτhτ+hτµ+hτe

ττbb→hh→H

Figure 8: Upper limits at 95% CL on the H ! hh ! bbtt cross section times branching fractionfor the µth (top left), eth (top right), thth (bottom left), and for final states combined (bottomright)

(X)→hh→bb𝝉𝝉: resultsCMS

)α-βcos(-1.0 -0.5 0.0 0.5 1.0

βta

n

123456789

10

95% CL Excluded:

Observed Expectedσ 1±

Expected Expectedσ 2±

= 300 GeVH = mA

2HDM type-II, m

CMS Unpublished (8 TeV)-119.7 fbττbb→hh→H

(GeV)Am250 300 350

βta

n

1.0

1.5

2.0

2.5

3.0

3.5

4.0

95% CL Excluded:

Observed Expectedσ 1±

Expected Expectedσ 2±

3 GeV± 125≠ hMSSMm

[260,350] GeV∉ HMSSM m

scenarioβMSSM low tan

CMSUnpublished

(8 TeV)-119.7 fbττbb→hh→H

(GeV)Am250 300 350

βta

n

1.0

1.5

2.0

2.5

3.0

3.5

4.0

95% CL Excluded:

Observed Expectedσ 1±

Expected Expectedσ 2±

3 GeV± 125≠ hMSSMm

[260,350] GeV∉ HMSSM m

scenarioβMSSM low tan

CMSUnpublished

(8 TeV)-119.7 fbττbb→hh→H

MSSM low tanβ scenario

(GeV)Am250 300 350

βta

n

1.0

1.5

2.0

2.5

3.0

3.5

4.0

95% CL Excluded:

Observed Expectedσ 1±

Expected Expectedσ 2±

3 GeV± 125≠ hMSSMm

[260,350] GeV∉ HMSSM m

scenarioβMSSM low tan

CMSUnpublished

(8 TeV)-119.7 fbττbb→hh→H

2HDM Type-II mH = mA = 300 GeV

The upper limit at 95% CL on the σ(ggH)×BR(H→hh→bb𝝉𝝉) is re-interpreted in 2HDM Type-II and MSSM models

”low tanβ” scenario: the value of MSUSY is increased until the mass of the lightest Higgs boson is consistent with 125 GeV over a range of low tan β and mA values

arXiv:1510.01181

Page 15: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 [GeV]Hm

300 400 500 600 700 800 900 1000

hh)

[pb]

→ B

R(H

×H

)→

(gg

σ

-110

1

10 Observed

Expected

expectedσ 1±

expectedσ 2±

ATLAS-1= 8 TeV, 20.3 fbs

)hadτ+bbehad

τµ (bbττ bb→hh

15

[GeV]ττm

0 100 200 300 400 500

Eve

nts

/ 2

5 G

eV

0

20

40

60

80

100

120

140

160

180

200

220

Data

Top quark

+jetsττ→Z

Others

τFake

Systematics

Non-reso. hh(10 pb)

ATLAS-1 = 8 TeV, 20.3 fbs

hadτ + ehad

τµ

(X)→hh→bb𝝉𝝉 arXiv:1509.04670ATLAS

Final states: e𝝉h μ𝝉h

divided in categories based on PT

𝝉𝝉 (< or > 100 GeV) number of b-jets (1 or ≥2)

Final discriminants used to extract the signal (1D analysis)non-resonant: m𝝉𝝉

Selection largely following the SM H→𝝉𝝉 analysis

Simulation

Embedded

“Fake-factor” method

[GeV]ττbbm

200 300 400 500 600 700 800 900

Eve

nts

/ 2

5 G

eV

0

10

20

30

40

50

60

70

Data

Top quark

+jetsττ→Z

Others

τFake

Systematics

H(300) hh(10 pb)

ATLAS-1 = 8 TeV, 20.3 fbs

hadτ + ehad

τµ

resonant: mbb𝝉𝝉

Table 2: The numbers of events predicted from background processes and observed in the data passing the finalselection of the resonant search for the four categories. The top quark background includes contributions from bothtt and the single top-quark production. The “others” background comprises diboson and Z! ee/µµ contributions.The numbers of events expected from the production of a mH = 300 GeV Higgs boson with a cross section of�(gg!H)⇥BR(H!hh) = 1 pb are also shown as illustrations. The uncertainties shown are the total uncertainties,combining statistical and systematic components.

nb = 1 nb � 2Process p⌧⌧T < 100 GeV p⌧⌧T > 100 GeV p⌧⌧T < 100 GeV p⌧⌧T > 100 GeVSM Higgs 0.5 ± 0.1 0.8 ± 0.1 0.1 ± 0.1 0.2 ± 0.1Top quark 30.3 ± 3.6 19.6 ± 2.5 30.9 ± 3.0 23.6 ± 2.5Z!⌧⌧ 38.1 ± 4.4 20.2 ± 3.7 6.8 ± 1.8 2.6 ± 1.0Fake ⌧had 37.0 ± 4.4 12.1 ± 1.7 13.7 ± 1.9 5.4 ± 1.0Others 3.2 ± 3.7 0.5 ± 1.5 0.7 ± 1.6 0.2 ± 0.7Total background 109.1 ± 8.6 53.1 ± 6.0 52.2 ± 8.2 32.1 ± 5.4Data 92 46 35 35

Signal mH = 300 GeV 0.8 ± 0.2 0.4 ± 0.2 1.5 ± 0.3 0.9 ± 0.2

[GeV]ττm

0 100 200 300 400 500

Eve

nts

/ 2

5 G

eV

0

20

40

60

80

100

120

140

160

180

200

220

Data

Top quark

+jetsττ→Z

Others

τFake

Systematics

Non-reso. hh(10 pb)

ATLAS-1 = 8 TeV, 20.3 fbs

hadτ + ehad

τµ

(a)

[GeV]ττbbm

200 300 400 500 600 700 800 900

Eve

nts

/ 2

5 G

eV

0

10

20

30

40

50

60

70

Data

Top quark

+jetsττ→Z

Others

τFake

Systematics

H(300) hh(10 pb)

ATLAS-1 = 8 TeV, 20.3 fbs

hadτ + ehad

τµ

(b)

Figure 3: Distributions of the final discriminants used to extract the signal: (a) m⌧⌧ for the nonresonant search and(b) mbb⌧⌧ for the resonant search. The top quark background includes contributions from both tt and the singletop-quark production. The background category labeled “Others” comprises diboson and Z!ee/µµ contributions.Contributions from single SM Higgs boson production are included in the background estimates, but are too smallto be visible on these distributions. As illustrations, the expected signal distributions assume a cross section of10 pb for Higgs boson pair production for both the nonresonant and resonant searches. In (b), a resonance massof 300 GeV is assumed. The gray hatched bands represent the uncertainties on the total backgrounds. Theseuncertainties are largely correlated from bin to bin.

12

Table 2: The numbers of events predicted from background processes and observed in the data passing the finalselection of the resonant search for the four categories. The top quark background includes contributions from bothtt and the single top-quark production. The “others” background comprises diboson and Z! ee/µµ contributions.The numbers of events expected from the production of a mH = 300 GeV Higgs boson with a cross section of�(gg!H)⇥BR(H!hh) = 1 pb are also shown as illustrations. The uncertainties shown are the total uncertainties,combining statistical and systematic components.

nb = 1 nb � 2Process p⌧⌧T < 100 GeV p⌧⌧T > 100 GeV p⌧⌧T < 100 GeV p⌧⌧T > 100 GeVSM Higgs 0.5 ± 0.1 0.8 ± 0.1 0.1 ± 0.1 0.2 ± 0.1Top quark 30.3 ± 3.6 19.6 ± 2.5 30.9 ± 3.0 23.6 ± 2.5Z!⌧⌧ 38.1 ± 4.4 20.2 ± 3.7 6.8 ± 1.8 2.6 ± 1.0Fake ⌧had 37.0 ± 4.4 12.1 ± 1.7 13.7 ± 1.9 5.4 ± 1.0Others 3.2 ± 3.7 0.5 ± 1.5 0.7 ± 1.6 0.2 ± 0.7Total background 109.1 ± 8.6 53.1 ± 6.0 52.2 ± 8.2 32.1 ± 5.4Data 92 46 35 35

Signal mH = 300 GeV 0.8 ± 0.2 0.4 ± 0.2 1.5 ± 0.3 0.9 ± 0.2

[GeV]ττm

0 100 200 300 400 500

Eve

nts

/ 2

5 G

eV

0

20

40

60

80

100

120

140

160

180

200

220

Data

Top quark

+jetsττ→Z

Others

τFake

Systematics

Non-reso. hh(10 pb)

ATLAS-1 = 8 TeV, 20.3 fbs

hadτ + ehad

τµ

(a)

[GeV]ττbbm

200 300 400 500 600 700 800 900

Eve

nts

/ 2

5 G

eV

0

10

20

30

40

50

60

70

Data

Top quark

+jetsττ→Z

Others

τFake

Systematics

H(300) hh(10 pb)

ATLAS-1 = 8 TeV, 20.3 fbs

hadτ + ehad

τµ

(b)

Figure 3: Distributions of the final discriminants used to extract the signal: (a) m⌧⌧ for the nonresonant search and(b) mbb⌧⌧ for the resonant search. The top quark background includes contributions from both tt and the singletop-quark production. The background category labeled “Others” comprises diboson and Z!ee/µµ contributions.Contributions from single SM Higgs boson production are included in the background estimates, but are too smallto be visible on these distributions. As illustrations, the expected signal distributions assume a cross section of10 pb for Higgs boson pair production for both the nonresonant and resonant searches. In (b), a resonance massof 300 GeV is assumed. The gray hatched bands represent the uncertainties on the total backgrounds. Theseuncertainties are largely correlated from bin to bin.

12

[GeV]Hm

300 400 500 600 700 800 900 1000

hh

) [p

b]

→ B

R(H

×H

)→

(gg

σ

-110

1

10 Observed

Expected

expectedσ 1±

expectedσ 2±

ATLAS-1= 8 TeV, 20.3 fbs

)hadτ+bbehad

τµ (bbττ bb→hh

Numbers of events predicted from background and observed in the data

Page 16: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

(X)→hh→bb𝜸𝜸

BRSM(hh→bb𝛾𝛾) ~ 0.26%

CMS-HIG-13-032 Phys. Rev. Lett. 114, 081802 (2015)

No

trev

iew

ed

,fo

rin

tern

al

circu

latio

no

nly

Draft version 1.0

ATLAS NOTEFebruary 26, 2013

Study of the spin of the Higgs-like particle in the H ! WW(⇤) ! e⌫µ⌫1

channel with 20.7 fb�1 of⇧

s = 8 TeV data collected with the ATLAS2

detector3

The ATLAS Collaboration4

Abstract5

Recently, the ATLAS collaboration reported the observation of a new neutral particle6

in the search for the Standard Model Higgs boson. The measured production rate of the7

new particle is consistent with the Standard Model Higgs boson with a mass of about 1258

GeV, but its other physics properties are unknown. Presently, the only constraint on the9

spin of this particle stems from the observed decay mode to two photons, which disfavours10

a spin-1 hypothesis. This note reports on the compatibility of the observed excess in the11

H ⌅ WW(⇥) ⌅ e⇥µ⇥ search arising from either a spin-0 or a spin-2 particle with positive12

charge-parity. Data collected in 2012 with the ATLAS detector favours a spin-0 signal, and13

results in the exclusion of a spin-2 signal at 95% confidence level if one assumes a qq ⌅ X14

production fraction larger than 25% for a spin-2 particle, and at 91% confidence level if one15

assumes pure gg production.16

c⇤ Copyright 2013 CERN for the benefit of the ATLAS Collaboration.Reproduction of this article or parts of it is allowed as specified in the CC-BY-3.0 license.

- John Alison - Experimental Studies of hh Higgs Coupling 2014

hh Decay

13

-810

-710

-610

-510

-410

-310

-210

-110

1

bb WW gg oo cc ZZ aa aZ µµ

µµ

aZ

aa

ZZ

cc

oo

ggWW

bb 33%

3e-3

7%

25%

Phenomenologically rich set of final states. hh-Br

- John Alison - Higgs Coupling 2014

Larger Br-h decay

Rarer Br-h decay

Page 17: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015

(X)→hh→bb𝜸𝜸

17

Follow SM h→𝛾𝛾 analysis no Primary Vertex efficiency issue thanks to the hadronic h

Three categories of events 0 b-tags: used as control region 1 and >=2 b-tags: medium and high purity region

Two ranges of resonances masses 260≤mX≤400 GeV : m𝛾𝛾 to extract the signal 400<mX≤1000 GeV : mkin𝛾𝛾jj to extract the signal

16 10 Conclusions

mX Observed limit (fb) Expected limit (fb)400 2.98 1.87450 1.76 1.42500 1.19 0.97550 1.45 0.80600 0.98 0.69650 0.61 0.60700 0.44 0.54800 0.31 0.46900 0.32 0.431000 0.33 0.431100 0.41 0.48

Table 6: Observed and median expected 95% CL limit from the high mass region: mX � 400GeV.

(GeV)X m300 400 500 600 700 800 900 1000 1100

) (fb

)bbγγ

→ H

H

→ B

R(X

×

X)

→(p

p σ -210

-110

1

10

210

WED: kl = 35, k/Mpl = 0.1, elementary top, no r/H mixing = 3 TeV)RΛradion ( = 1 TeV)RΛradion (

RS1 KK-graviton Bulk KK-graviton

Observed 95% upper limitExpected 95% upper limit

σ 1 ±Expected limit σ 2 ±Expected limit

= 8 TeVs -1CMS Preliminary L = 19.7 fb

Figure 8: Expected 95% CL upper limits on the cross section times branching ratios s(pp !X) ⇥ BR(X ! HH ! ggbb). Theory lines corresponding to WED models with radion, RS1KK-graviton and bulk KK-graviton are also shown. The results are obtained using the asymp-totic CLs approach. The vertical dashed line shows the separation between the low mass anal-ysis and high mass analysis. The limits for mX = 400 GeV are shown with both methods.

CMS CMS-PAS-HIG-13-032

(GeV)γγm100 110 120 130 140 150 160 170 180

Even

ts /

( 1 G

eV )

1

2

3

4

5

6

= 8 TeV s -1 CMS Preliminary L = 19.7 fb

= 300 GeVX mData Fit

σ 2 ±Fit σ 1 ±Fit

High puritybbγγ → HH →X

(GeV)kin jjγγm

400 500 600 700 800 900 1000 1100 1200

Even

ts /

( 10

GeV

)0

1

2

3

4

5

6

7

8

9 = 8 TeV s -1 CMS Preliminary L = 19.7 fb

Data Fit

σ 2 ±Fit σ 1 ±Fit

Medium puritybbγγ → HH →X

Page 18: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015

4

the signal plus background fit of the non-resonance anal-ysis, and in the resonance search it is transformed intoa 1.7% uncertainty on the number of signal events inthe mass window. The uncertainty for the acceptanceof the m

��

cuts on non-Higgs boson backgrounds is esti-mated by comparing fits of m

��

to data in control regionswith reversed photon identification or b-tagging require-ments, and using di↵erent functional forms. The largestdeviation observed from these fits (11%) is used for allsearches.

Three components contribute to the uncertainty on✏m��bb

, and are combined in quadrature. (1) The lim-ited number of events in the control region with fewerthan two b-tags used for the Landau fit lead to a relativestatistical uncertainty of 3–18% that varies as a functionof m

X

. (2) The m��jj

shape for untagged jets might notexactly mirror the one for tagged jets. The tagged anduntagged samples are compared in simulation and the rel-ative di↵erence in ✏

m��bbis taken as the uncertainty. This

value varies with mX

and is always less than 30%. (3)Finally, an uncertainty of 16–30%, depending on m

X

, isincluded to cover the choice of the analytic function. Thiswas evaluated via comparisons of Landau shapes to al-ternate functions in simulation, including Landau shapeswhere the width varies withm

��bb

, as well as Crystal Ballfunctions. Potential contamination from single Higgs bo-son processes in the control region is estimated to be lessthan 4% and is subtracted with negligible impact on theshape.

Uncertainties due to the b-tagging calibration are typ-ically 2–4% for both the single Higgs boson and signalprocesses. Uncertainties due to the jet energy scale are7% (22%) for single Higgs boson backgrounds in the non-resonance (resonant) analysis, and 1.4% (4.4%) for signalprocesses. Uncertainties due to the jet energy resolutionare 4.8% (21%) for single Higgs boson backgrounds, and6.3% (9.3%) for signal processes. The uncertainty onthe integrated luminosity is 2.8%. It is derived, follow-ing the same methodology as that detailed in Ref. [34],from a preliminary calibration of the luminosity scale de-rived from beam-separation scans performed in Novem-ber 2012.

The combined signal plus background fit for the non-resonance analysis is shown in Fig. 1. Within a ±2�

m��

window around the Higgs boson mass, 1.5 events are ex-pected, with 1.3 ± 0.5 from the continuum backgroundand 0.17 ± 0.04 from single Higgs boson production,which is dominated by tth events. About 0.04 eventsare expected from SM Higgs boson pair production.Five events are observed, corresponding to 2.4� fromthe background-only hypothesis. The 95% confidencelevel (CL) upper limit on the Higgs boson pair produc-tion cross section is calculated using the frequentist CL

S

method [51]. Exclusions and significances are evaluatedusing pseudo-experiments. Assuming SM branching ra-tios for the light Higgs boson decays, the expected upperlimit is 1.0+0.5

�0.2 pb; the observed limit is 2.2 pb.For the resonance analysis, as before, SM branching

[GeV]Xm300 350 400 450 500

hh) [

pb]

→ B

R(X

× Xσ

0.5

1

1.5

2

2.5

3

3.5

4

4.5ATLAS∫ = 8 TeVs at -1Ldt = 20 fb

Observed 95% CL Limitσ1±Expected Limit σ2±Expected Limit

Type I 2HDM:)=-0.05α-β=1, cos(β tan

FIG. 3. A 95% CL upper limit on the cross section timesbranching ratio of a narrow resonance decaying to pairs ofHiggs bosons as a function of mX (see text for more details).

fractions for the light Higgs boson are assumed. The ex-pected exclusion improves from 1.7 to 0.7 pb as a func-tion ofm

X

from 260 to 500 GeV, as shown in Fig. 3. Thisbehavior derives from increased event-level acceptance atlarger masses. The observed exclusion ranges from 3.5to 0.7 pb. The five events selected in the m

��

signalregion are shown in m

��bb

, in Fig. 2. The local proba-bility of compatibility to the background-only hypothe-sis, p0, reaches a minimum of 0.002 at m

X

= 300 GeV,corresponding to 3.0�. After accounting for the look-elsewhere e↵ect [52, 53], the global probability of suchan excess occurring at any mass in the range studied is0.019, corresponding to 2.1�. The number of events ly-ing within the m

��bb

window of each mass hypothesis isreadily apparent in ‘steps’ in the exclusion plot.The limits derived are juxtaposed in Fig. 3 with the

expectation from a sample type I 2HDM [32, 33, 54] notexcluded by current data with cos(� � ↵) = �0.05 andtan(�) = 1. The heavy Higgs bosons are taken to bedegenerate in mass, and the mass of the lightest CP-even Higgs boson is set to 125 GeV. All major produc-tion mechanisms of H ! hh are considered. Cross sec-tions and branching ratios were calculated as describedin Ref. [55].In conclusion, this Letter presents searches for reso-

nant and non-resonant Higgs boson pair production using20.3 fb-1 of proton–proton collision data at

ps = 8 TeV

generated by the Large Hadron Collider and recorded bythe ATLAS detector in 2012. A 95% confidence level up-per limit is placed on the non-resonant production crosssection at 2.2 pb, while the expected limit is 1.0+0.5

�0.2 pb.The di↵erence derives from a small excess of events, cor-responding to 2.4�.In the search for a narrow resonance decaying to a pair

of Higgs bosons, the expected exclusion on the produc-tion cross section falls from 1.7 pb for a resonance at260 GeV to 0.7 pb at 500 GeV. The observed exclusionranges from 0.7–3.5 pb. It is weaker than expected for

18

3

experiment instead of a simultaneous fit. The m��

reso-lution, �

m�� , is set to the expected value of 1.6 GeV, andthe diphoton mass is required to be within ±2�

m�� of theHiggs boson mass, m

h

= 125.5GeV [3]. The acceptanceof this requirement on background events without Higgsbosons, ✏

m�� , is measured by fitting an exponential func-tion to the m

��

sidebands for events with fewer than twob-tagged jets. For this fit, them

��

region ofmh

±5 GeV isexcluded to eliminate any potential contamination fromresonant Higgs boson production. For N observed eventswith two b-tags in the sideband (|m

��

�mh

| > 2�m�� ),

the number of expected non-Higgs boson backgroundevents (N

m�� ) within 2�m�� around m

h

is given by:

Nm�� = N

✏m��

1� ✏m��

, (1)

where the denominator compensates for the fact that✏m�� = 0.13 is derived relative to the full m

��

spectrumwhile N contains only those events in the sidebands.

Before reconstructing the four-object mass, m��bb

,a scaling factor of m

h

/mbb

is applied to the four-momentum of the bb system, where m

h

is set to the valueof 125 GeV used in simulation. This improves the m

��bb

resolution by 30–60% depending on the mass hypothesis,without biasing or significantly altering the shape of thebackground. Requirements are then made on m

��bb

, toselect the smallest window containing 95% of the signalevents in the narrow-width simulation. These require-ments vary linearly with the mass, m

X

, of the resonanceconsidered. The width of the signal window varies from17 GeV at m

X

= 260 GeV to 60 GeV at mX

= 500 GeV.The acceptance for the continuum background to passthis requirement, ✏

m��bb, also varies with m

X

. It is mea-sured using events in data with |m

��

�mh

| < 2�m�� and

fewer than two b-tags. Studies in both data sidebandsand simulation show that the shapes of m

��bb

and m��jj

agree within statistical uncertainties. The distributionof m

��jj

in data is fitted with a Landau function, whichis integrated in the signal window to obtain ✏

m��bbfor

each mass hypothesis. The bottom panel of Fig. 2 showsthis fit. The value of ✏

m��bbis small (< 8%) at low and

high mX

, and peaks at 18% for mX

= 300 GeV. Thecombined acceptance and selection e�ciency for a reso-nance signal to pass all requirements varies from 3.8% atm

X

= 260 GeV to 8.2% at mX

= 500 GeV.The total background from sources without Higgs bo-

son decays in the resonance analysis NB is given by:

NB = N✏m��

1� ✏m��

✏m��bb

, (2)

where NB and ✏m��bb

are functions of mX

. Uncertaintieson this extrapolation are described below.

Because they are not accounted for by the abovem

��

sideband techniques, contributions from single Higgsbosons produced in association with jets (particularlywith cc or bb pairs) are estimated using simulation. In

[GeV]bbγγConstrained m

Even

ts /

5 G

eV

-210

-110

1

10 Signal RegionDataControl Region FitSingle Higgs Boson

=1 pbhhBR×Xσ=300 GeV, Xm

ATLAS

∫ = 8 TeVs, -1Ldt = 20 fb

[GeV]jjγγConstrained m200 300 400 500 600 700 800 900

Even

ts /

20 G

eV

1

10< 2 b-Tag Control Region

DataLandau Fit

FIG. 2. (Upper plot) The constrained four-object invari-ant mass, m��jj , for data events in the resonance signal re-gion. The expected backgrounds are also shown. A narrowwidth resonance at 300 GeV is displayed for comparison only.(Lower plot) The diphoton invariant mass spectrum in thecontinuum background from events with fewer than two b-tags and the corresponding fitted curve, the shape of whichis also used in the upper plot.

the resonance analysis, the yield from the non-resonantSM hh processes is similarly included. SM cross sectionsand branching fractions are assumed in all cases [29].Most systematic uncertainties are small when com-

pared to statistical uncertainties, in particular for theresonance search.The evaluation of experimental uncertainties on pho-

ton identification (2.4%) and isolation e�ciencies (2%)follow the methods used in the inclusive ATLAS h ! ��analyses [3, 29, 38]. The theoretical uncertainties on thesingle Higgs boson backgrounds are similarly adopted.Uncertainties from the modeling of Higgs bosons pro-duced in association with extra heavy-flavor quarks areestimated from comparisons of data to simulation pre-dictions for both tt [49] and W boson [50] productionwith extra heavy-flavor jets. The tt (W boson) studyprovides the uncertainty on extra heavy-flavor contentin gluon- (quark-)initiated final states, leading to a 14%uncertainty on the combined single Higgs boson contri-bution in the signal region. PDF and scale uncertaintieson SM hh production are taken from Ref. [8].Because of the cuts on the ratio pT/m��

, photon en-ergy scale uncertainties are negligible. The uncertaintyof 13% on the diphoton mass resolution in the non-resonance search is implemented in the resonance anal-ysis as a 1.6% uncertainty on the number of events mi-grating into and out of the signal region. This representsthe fraction of events where an upward variation of thephoton resolution causes the diphoton mass to leave them

h

± 2�m�� window required for the signal region. The

uncertainty on mh

impacts the peak position in m��

in

Limit non-resonant analysis: Expected: 1.0+0.2

-0.5 pb Observed: 2.2 pb (2.4σ)

Limit resonant analysis: 1pb in the higher mass region Local excesses below 350 GeV around 3.5 pb

(X)→hh→bb𝜸𝜸 ATLAS Phys. Rev. Lett. 114, 081802 (2015)

Follow SM h→𝛾𝛾 measurement analysis

Only one category of events

Search for: → spin 0 resonances in the 260≤mX≤500 GeV mass range SM Higgs boson → non-resonant pair production

Constrained m𝛾𝛾jj to extract the signal

Page 19: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Other channels

Phys. Rev. D 90 (2014) 112013 Phys. Rev. Lett. 114, 081802 (2015)

Multileptons Lepton plus photons (X)→hh→WW𝜸𝜸

BRSM(hh→WWWW) ~ 4.6%

BRSM(hh→WWττ) ~ 2.7%

BRSM(hh→WW𝛾𝛾) ~ 0.1%

BRSM(hh→WW𝛾𝛾) ~ 0.1%

… … …. … … ….

No

trev

iew

ed

,fo

rin

tern

al

circu

latio

no

nly

Draft version 1.0

ATLAS NOTEFebruary 26, 2013

Study of the spin of the Higgs-like particle in the H ! WW(⇤) ! e⌫µ⌫1

channel with 20.7 fb�1 of⇧

s = 8 TeV data collected with the ATLAS2

detector3

The ATLAS Collaboration4

Abstract5

Recently, the ATLAS collaboration reported the observation of a new neutral particle6

in the search for the Standard Model Higgs boson. The measured production rate of the7

new particle is consistent with the Standard Model Higgs boson with a mass of about 1258

GeV, but its other physics properties are unknown. Presently, the only constraint on the9

spin of this particle stems from the observed decay mode to two photons, which disfavours10

a spin-1 hypothesis. This note reports on the compatibility of the observed excess in the11

H ⌅ WW(⇥) ⌅ e⇥µ⇥ search arising from either a spin-0 or a spin-2 particle with positive12

charge-parity. Data collected in 2012 with the ATLAS detector favours a spin-0 signal, and13

results in the exclusion of a spin-2 signal at 95% confidence level if one assumes a qq ⌅ X14

production fraction larger than 25% for a spin-2 particle, and at 91% confidence level if one15

assumes pure gg production.16

c⇤ Copyright 2013 CERN for the benefit of the ATLAS Collaboration.Reproduction of this article or parts of it is allowed as specified in the CC-BY-3.0 license.

- John Alison - Experimental Studies of hh Higgs Coupling 2014

hh Decay

13

-810

-710

-610

-510

-410

-310

-210

-110

1

bb WW gg oo cc ZZ aa aZ µµ

µµ

aZ

aa

ZZ

cc

oo

ggWW

bb 33%

3e-3

7%

25%

Phenomenologically rich set of final states. hh-Br

- John Alison - Higgs Coupling 2014

Larger Br-h decay

Rarer Br-h decay

Page 20: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015

(GeV)missTE

0-30 30-50 50-100 >100

Even

ts/B

in

-110

1

10Data

hh 300 GeV→HBkg. uncertaintiesMisidentifiedttWZZZWttZtt

SM Higgs boson

CMS (8 TeV)-119.5 fb4-leptons: OSSF1, off-Z, no taus, no b-jets

Multileptons / lepton plus photons

20

CMS Phys. Rev. D 90 (2014) 112013

>=3 leptons covering 7 final states: WWWW, WWZZ, WWττ, ZZZZ, ZZττ, ZZbb, and ττττ

2 photons plus >=1 lepton covering 3 final states: WW𝛾𝛾, ZZ𝛾𝛾, and ττ𝛾𝛾

)α-βcos(-0.6 -0.4 -0.2 0 0.2 0.4 0.6

βta

n

-110

1

10

CMS (8 TeV)-119.5 fb = 300 GeV

H hh, m→Type II 2HDM H

Observed 95% CL limitsExpected 95% CL limits

σ1 ±Expected σ2 ±Expected

)α-βcos(-0.6 -0.4 -0.2 0 0.2 0.4 0.6

βta

n

-110

1

10

CMS (8 TeV)-119.5 fb = 300 GeV

H hh, m→Type I 2HDM H

Observed 95% CL limitsExpected 95% CL limits

σ1 ±Expected σ2 ±Expected

2HDM Type-I mH = 300 GeV 2HDM Type-II mH = 300 GeV

Page 21: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/201521

(X)→hh→WW𝜸𝜸 ATLAS arXiv:1509.04670

𝛾𝛾 side following SM h→𝛾𝛾 analysis used to tag the Higgs bosonWW side additional lepton+jets (b-tag veto)

[GeV]H m

250 300 350 400 450 500

hh)

[pb]

→B

R(H

×H

)→

(gg

σ

5

10

15

20

25

30

35

Observed

Expected

expectedσ 1±

expectedσ 2±

ATLAS = 8 TeV,s -120.3 fb

qq)νlγγWW (γγ →hh

[GeV] γγm

110 120 130 140 150 160

Eve

nts

/ 2

.5 G

eV

0

1

2

3

4

5

6

7

8

9

Data

Continuum background

SM Higgs contribution

-1 = 8 TeV, 20.3 fbsATLAS

[GeV] γγm

110 120 130 140 150 160

Eve

nts

/ 2

.5 G

eV

200

400

600

800

1000

1200

Data (control region)

Continuum background

SM Higgs contribution

-1 = 8 TeV, 20.3 fbsATLAS

Relaxed requirements Final selection

All final selections except on the lepton and EmissT

13 events

Selection efficiency → SM non-resonant 2.9%. → resonant production, from 1.7% at 260 GeV to 3.3% at 500 GeV

SM non-resonant p-value of the background-only hypothesis is 3.8% (1.8σ)

Page 22: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Combination

Page 23: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015

Upper limits

23

[GeV]H m

300 400 500 600 700 800 900 1000

hh

) [p

b]

→B

R(H

×H

)→

(gg

σ

-210

-110

1

10

210

expττbb

expγγWW

expγγbb

bbbb exp

Observed

Expected

expectedσ 1±

expectedσ 2±

ATLASATLAS-1 = 8 TeV, 20.3 fbs

arXiv:1509.04670ATLAS

values of the Higgs boson mass mh and, therefore, have been updated using a common mass value ofmh = 125.4 GeV [24] for the combinations. The decay branching ratios of the Higgs boson h and theiruncertainties used in the combinations are taken from Ref. [27]. Table 3 is a summary of the number ofcategories and final discriminants used for each analysis.

Table 3: An overview of the number of categories and final discriminant distributions used for both the nonresonantand resonant searches. Shown in the last column are the mass ranges of the resonant searches.

hh Nonresonant search Resonant searchfinal state Categories Discriminant Categories Discriminant mH [GeV]��bb 1 m�� 1 event yields 260–500��WW⇤ 1 event yields 1 event yields 260–500

bb⌧⌧ 4 m⌧⌧ 4 mbb⌧⌧ 260–1000bbbb 1 event yields 1 mbbbb 500–1500

The four individual analyses are sensitive to di↵erent kinematic regions of the hh production and decays.The combination is performed assuming that the relative contributions of these regions to the total crosssection are modeled by the MadGraph5 [39] program used to simulate the hh production.

9 Results

In this section, the limits on the nonresonant and resonant searches are derived. The results of the hh!bb⌧⌧ and hh!��WW⇤ analyses are first determined and then combined with previously published resultsof the hh! ��bb and hh! bbbb analyses. The impact of the leading systematic uncertainties is alsodiscussed.

The observed and expected upper limits at 95% CL on the cross section of nonresonant production ofa Higgs boson pair are shown in Table 4. These limits are to be compared with the SM prediction of9.9 ± 1.3 fb [17] for gg!hh production with mh = 125.4 GeV. Only the gluon fusion production processis considered. The observed (expected) cross-section limits are 1.6 (1.3) pb and 11.4 (6.7) pb from thehh! bb⌧⌧ and hh!��WW⇤ analyses, respectively. Also shown in the table are the cross-section limitsrelative to the SM expectation. The results are combined with those of the hh! ��bb and hh! bbbbanalyses. The p-value of compatibility of the combination with the SM hypothesis is 4.4%, equivalent to1.7 standard deviations. The low p-value is a result of the excess of events observed in the hh! ��bbanalysis. The combined observed (expected) upper limit on �(gg!hh) is 0.69 (0.47) pb, correspondingto 70 (48) times the cross section predicted by the SM. The hh! bbbb analysis has the best expectedsensitivity followed by the hh!��bb analysis. The observed combined limit is slightly weaker than thatof the hh!bbbb analysis, largely due to the aforementioned excess.

The impact of systematic uncertainties on the cross-section limits is studied using the signal-strengthparameter µ, defined as the ratio of the extracted to the assumed signal cross section (times branching ratioBR(H!hh) for the resonant search). The resulting shifts in µ depend on the actual signal-strength value.For illustration, they are evaluated using a cross section of 1 pb for gg! (H!)hh, comparable to the limitsset. The e↵ects of the most important uncertainty sources are shown in Table 5. The leading contributionsare from the background modeling, b-tagging, the h decay branching ratios, jet and Emiss

T measurements.

16

Table 4: The expected and observed 95% CL upper limits on the cross sections of nonresonant gg!hh productionatp

s = 8 TeV from individual analyses and their combinations. SM values are assumed for the h decay branchingratios. The cross-section limits normalized to the SM value are also included.

Analysis ��bb ��WW⇤ bb⌧⌧ bbbb Combined

Upper limit on the cross section [pb]Expected 1.0 6.7 1.3 0.62 0.47Observed 2.2 11 1.6 0.62 0.69

Upper limit on the cross section relative to the SM predictionExpected 100 680 130 63 48Observed 220 1150 160 63 70

Table 5: The impact of the leading systematic uncertainties on the signal-strength parameter µ of a hypothesizedsignal for both the nonresonant and resonant (mH = 300, 600 GeV) searches. For the signal hypothesis, a Higgsboson pair production cross section (�(gg!hh) or �(gg!H) ⇥ BR(H!hh)) of 1 pb is assumed.

Nonresonant search Resonant searchmH = 300 GeV mH = 600 GeV

Source �µ/µ [%] Source �µ/µ [%] Source �µ/µ [%]Background model 11 Background model 15 b-tagging 10b-tagging 7.9 Jet and Emiss

T 9.9 h BR 6.3h BR 5.8 Lepton and ⌧had 6.9 Jet and Emiss

T 5.5Jet and Emiss

T 5.5 h BR 5.9 Luminosity 2.7Luminosity 3.0 Luminosity 4.0 Background model 2.4Total 16 Total 21 Total 14

The large impact of the b-tagging systematic uncertainty reflects the relatively large weight of the hh!bbbb analysis in the combination.

For the resonant production, limits are set on the cross section of gg!H production of the heavy Higgsboson times its branching ratio BR(H ! hh) as a function of the heavy Higgs boson mass mH . Theobserved (expected) limits of the hh!bb⌧⌧ and hh!��WW⇤ analyses are illustrated in Fig. 5 and listedin Table 6 (along with results from the hh! ��bb and hh! bbbb analyses). The mH search ranges are260–1000 GeV for hh! bb⌧⌧ and 260–500 GeV for hh! ��WW⇤. For the hh! bb⌧⌧ analysis, theobserved limit around mH ⇠ 300 GeV is considerably lower than the expectation, reflecting the deficitin the observed mbb⌧⌧ distribution. At high mass, the limits are correlated since a single bin is usedfor mbb⌧⌧ & 400 GeV. The decrease in the limit as mH increases is a direct consequence of increasingselection e�ciency for the signal. This is also true for the hh!��WW⇤ analysis as the event selection isindependent of mH .

The hh ! ��bb and hh ! bbbb analyses are published separately and the mass range covered by thetwo analyses are 260–500 GeV and 500–1500 GeV, respectively. The results of these four analyses,summarized in Table 6, are combined for the mass range 260–1000 GeV assuming the SM values of theh decay branching ratios. To reflect the better mass resolutions of the hh!bbbb and hh!��bb analyses,

17

The four individual analyses are sensitive to different kinematic regions of the hh production and decays

non-resonant

resonant

Page 24: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015

[GeV]Am220 240 260 280 300 320 340 360 380 400

βta

n1

1.21.41.61.8

22.22.4

250

275

300

325

350

375

low-tb-high-1 = 8 TeV, 20.3 fbs ATLASObserved exclusion Expected exclusion

(GeV)HConstant m expectedσ 1 ± < 122 GeVhm

Interpretation

24

arXiv:1509.04670ATLAS

[GeV]Am220 240 260 280 300 320 340 360 380 400

βta

n

11.21.41.61.8

22.22.4

250

275

300

325

350

375

hMSSM-1 = 8 TeV, 20.3 fbs ATLASObserved exclusion Expected exclusion

(GeV)HConstant m expectedσ 1 ±

Light CP-even Higgs boson is fixed to 125 GeV in the whole parameter space

Light CP-even Higgs boson mass is not fixed. Gray shaded region the mass of h is inconsistent with the measured value.

The upper limits on σ(gg→H)×BR(H→hh) can be interpreted as exclusion regions in the (tanβ, mA) plane.

Page 25: Higgs boson self-interactions in SM and BSM (ATLAS+CMS)...Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015 Introduction 3 Double Higgs boson (hh) production

Roberto Salerno (LLR) - Higgs Coupling 2015 - Lumley Castle - 13/10/2015

Summary

25

Search for hh final state in Run1 performed by both LHC experiments investigating a large variety of final states

The non-resonant search is far from SM sensitivity (50x SM) but new physics can be probed

Limits on resonant hh set on wide mass range