Fitting Higgs couplings in an EFT approach Oscar boli Universidade - - PowerPoint PPT Presentation

Fitting Higgs couplings in an EFT approach

Oscar Éboli Universidade de São Paulo

HEFT-2015

in collaboration with Corbett, Gonçalves, Gonzalez-Fraile, Gonzalez-Garcia, Plehn, Rauch

Goal: comprehensive analysis of couplings of the Higgs

• assume a narrow CP-even scalar
• use SFitter to fit available LHC data:

production/decay mode ATLAS CMS H → WW X X H → ZZ X X H → γγ X X H → τ ¯ τ X X H → b¯ b X X H → Zγ X X H → invisible X X t¯ tH production X X kinematic distributions X

• ff-shell rate

X X

• frequentist likelihood everywhere
• SFitter is flexible to study theoretical uncertainties

[flat distribution ; uncorrelated uncertainties for production] [159] [14] [37]

[SFitter: Gonzalez-Fraile, Klute, Plehn, Rauch, Zerwas]

1.How well does the SM describe the Higgs data?

[Corbett, OE, Gonçalves, Gonzalez-Fraile, Plehn, Rauch: arXiv:1505.05516]

• assume SM operators with free couplings:[nonlinear sigma model: Alonso et al.; Buchalla et al.; Brivio et al,…]

L = LSM + ∆W gmW H W µWµ + ∆Z g 2cw mZH ZµZµ − X

τ,b,t

∆f mf v H ¯ fRfL + h.c.

• + ∆gFG

H v GµνGµν + ∆γFA H v AµνAµν + invisible decays

corresponding changes in the Higgs couplings:

gx = gSM

x

(1 + ∆x) gγ = gSM

γ

(1 + ∆SM

γ

+ ∆γ) ≡ gSM

γ

(1 + ∆SM+NP

γ

) gg = gSM

g

(1 + ∆SM

g

+ ∆g) ≡ gSM

g

(1 + ∆SM+NP

g

)

tree level new physics

[keeping values for finite loop masses in calculations]

• flips the sign of the SM couplings

∆x = −2

• only rate measurements (of course!)

• presenting the SM like solutions with
• slowly increasing the number of free parameters
• equal tree level deviations

∆g = ∆γ = 0 ∆H

• 0.6
• 0.4
• 0.2

0.2 0.4 ∆H ∆V ∆f ∆W ∆Z ∆t ∆b ∆τ ∆Z/W ∆b/τ ∆b/W

gx = gx

SM (1+∆x) L=4.5-5.1(7 TeV)+19.4-20.3(8 TeV) fb-1, 68% CL: ATLAS + CMS

SM exp. data

injected SM Higgs signal controls hgg coupling (expected 15% error)

• adding new loop contributions (7 parameter fit)

∆g ∆γ

• 0.8
• 0.6
• 0.4
• 0.2

0.2 0.4 0.6 0.8 ∆

W

∆

Z

∆

t

∆

b

∆

τ

∆

γ

∆

g

∆

γ

S M + N P

∆

g

S M + N P

∆

Z / W

∆

b / τ

∆

b / W

gx = gx

SM (1+∆x) L=4.5-5.1(7 TeV)+19.4-20.3(8 TeV) fb-1, 68% CL: ATLAS + CMS

SM exp. data

effect of ttH (expected 30% error)

slightly decrease

• analogously < 0.7 (1.8) at 68% (95%) CL

∆γZ

∆g ∆t

• 4
• 3
• 2
• 1

1 2 3 4

• 4
• 3
• 2
• 1

1 2 ∆g ∆t

• 4
• 3
• 2
• 1

1 2 3 4

• 4
• 3
• 2
• 1

1 2 ∆γ ∆g

• 4
• 3
• 2
• 1

1

• 4
• 3
• 2
• 1

1 2 3 4 ∆γ ∆g

• 4
• 3
• 2
• 1

1

• 4
• 3
• 2
• 1

1 2 3 4 ∆(-2 ln L)

2 4 6 8 10 12 14 16

∆t

• multiple solutions due to degeneracy
• and contribute to gluon fusion production

∆g ∆x = 0 ⇐ ⇒ ∆x = −2 ∆W > −1 7 parameter fit plays indirect role ∆t

• 4
• 3
• 2
• 1

1 2 ∆b

• 0.6
• 0.4
• 0.2

0.2 0.4 0.6 ∆g

SM+NP

• 4
• 3
• 2
• 1

1 2 ∆b

• 0.6
• 0.4
• 0.2

0.2 0.4 0.6 ∆g

SM+NP

• 4
• 3
• 2
• 1

1 2 ∆b

• 0.4
• 0.2

0.2 0.4 0.6 0.8 1 ∆γ

SM+NP

• 4
• 3
• 2
• 1

1 2 ∆b

• 0.4
• 0.2

0.2 0.4 0.6 0.8 1 ∆γ

SM+NP

• further interesting correlations due to σ(pp → h → γγ)

∆t and ∆g ∆t ∆W and ∆γ

σon-shell

i→H→f ∝

g2

i (mH) g2 f(mH)

ΓH

• adding invisible decays (8 parameter fit)

BRinv < 0.31 at 95% CL

• 0.8
• 0.6
• 0.4
• 0.2

0.2 0.4 0.6 0.8 ∆W ∆Z ∆t ∆b ∆τ ∆γ ∆g BRinv ∆γ

SM+NP

∆g

SM+NP

∆Z/W ∆b/τ ∆b/W

gx = gx

SM (1+∆x) L=4.5-5.1(7 TeV)+19.4-20.3(8 TeV) fb-1, 68% CL: ATLAS + CMS

SM exp. data

• there is no significant deviations from the SM predictions

∆W BRinv

• 0.6 -0.4 -0.2

0.2 0.4 0.6 0.1 0.2 0.3 0.4 0.5 0.6 ∆W BRinv

• 0.6 -0.4 -0.2

0.2 0.4 0.6 0.1 0.2 0.3 0.4 0.5 0.6

no correlation with ∆W,Z

• minor upward shift of all couplings due to correlation with total width

σon-shell

i→H→f ∝

g2

i (mH) g2 f(mH)

ΓH

details of the theoretical uncertainty treatment

• flat vs gaussian distributions for the theoretical uncertainties
• 0.5

0.5 2 4 6 8 10 12

SM+NP g

∆ (-2 ln L) ∆

SM exp. data data (Gauss)

injected SM rates induced th. uncertainties

• presently statistical errors dominate
• gaussian distributions lead to slightly larger 68% CL bands

• 0.8
• 0.6
• 0.4
• 0.2

0.2 0.4 0.6 0.8 ∆W ∆Z ∆t ∆b ∆τ ∆γ ∆g ∆γ

SM+NP

∆g

SM+NP

∆Z/W ∆b/τ ∆b/W

gx = gx

SM (1+∆x) L=4.5-5.1(7 TeV)+19.4-20.3(8 TeV) fb-1, 68% CL: ATLAS + CMS

SM exp. SM exp. (corr.) data data (corr.)

• correlated vs uncorrelated uncertainties (7 parameter fit):
• correlated uncertainties lead to slightly smaller errors

• 2. Linear effective lagrangians to describe the LHC data?
• new state belongs to SU(2) doublet
• consider SU(2) x U(1) invariant dimension-6 lagrangian

Leff = LSM + X

n

fn Λ2 On + · · ·

• There are 59 “independent” dimension-six operators
• our choice for the boson operators is

[Buchmuller & Wyler; Grzadkowski] [Corbett, OE, Gonzalez-Fraile, Gonzalez-Garcia]

OGG = Φ†Φ Ga

µνGaµν

OW W = Φ† ˆ Wµν ˆ W µνΦ OBB = Φ† ˆ Bµν ˆ BµνΦ OBW = Φ† ˆ Bµν ˆ W µνΦ OW = (DµΦ)† ˆ W µν(DνΦ) OB = (DµΦ)† ˆ Bµν(DνΦ) OΦ,1 = (DµΦ)† Φ Φ† (DµΦ) OΦ,2 = 1 2∂µ Φ†Φ

• ∂µ
• Φ†Φ
• OΦ,4 = (DµΦ)† (DµΦ)
• Φ†Φ
• DµΦ =
• ∂µ + ig0Bµ/2 + igσaW a

µ/2

• Φ

ˆ Bµν = ig0Bµν/2 ˆ Wµν = igσaW a

µν/2

with

[Hagiwara, Ishihara, Szalapski, Zeppenfeld]

• 2. Linear effective lagrangians to describe the LHC data?
• new state belongs to SU(2) doublet
• consider SU(2) x U(1) invariant dimension-6 lagrangian

Leff = LSM + X

n

fn Λ2 On + · · ·

• There are 59 “independent” dimension-six operators
• our choice for the boson operators is

[Buchmuller & Wyler; Grzadkowski] [Corbett, OE, Gonzalez-Fraile, Gonzalez-Garcia]

OGG = Φ†Φ Ga

µνGaµν

OW W = Φ† ˆ Wµν ˆ W µνΦ OBB = Φ† ˆ Bµν ˆ BµνΦ OBW = Φ† ˆ Bµν ˆ W µνΦ OW = (DµΦ)† ˆ W µν(DνΦ) OB = (DµΦ)† ˆ Bµν(DνΦ) OΦ,1 = (DµΦ)† Φ Φ† (DµΦ) OΦ,2 = 1 2∂µ Φ†Φ

• ∂µ
• Φ†Φ
• OΦ,4 = (DµΦ)† (DµΦ)
• Φ†Φ
• DµΦ =
• ∂µ + ig0Bµ/2 + igσaW a

µ/2

• Φ

ˆ Bµν = ig0Bµν/2 ˆ Wµν = igσaW a

µν/2

with

eliminated with EOM

[Hagiwara, Ishihara, Szalapski, Zeppenfeld]

• 2. Linear effective lagrangians to describe the LHC data?
• new state belongs to SU(2) doublet
• consider SU(2) x U(1) invariant dimension-6 lagrangian

Leff = LSM + X

n

fn Λ2 On + · · ·

• There are 59 “independent” dimension-six operators
• our choice for the boson operators is

[Buchmuller & Wyler; Grzadkowski] [Corbett, OE, Gonzalez-Fraile, Gonzalez-Garcia]

OGG = Φ†Φ Ga

µνGaµν

OW W = Φ† ˆ Wµν ˆ W µνΦ OBB = Φ† ˆ Bµν ˆ BµνΦ OBW = Φ† ˆ Bµν ˆ W µνΦ OW = (DµΦ)† ˆ W µν(DνΦ) OB = (DµΦ)† ˆ Bµν(DνΦ) OΦ,1 = (DµΦ)† Φ Φ† (DµΦ) OΦ,2 = 1 2∂µ Φ†Φ

• ∂µ
• Φ†Φ
• OΦ,4 = (DµΦ)† (DµΦ)
• Φ†Φ
• constrained by EWPD

DµΦ =

• ∂µ + ig0Bµ/2 + igσaW a

µ/2

• Φ

ˆ Bµν = ig0Bµν/2 ˆ Wµν = igσaW a

µν/2

with

eliminated with EOM

[Hagiwara, Ishihara, Szalapski, Zeppenfeld]

• 2. Linear effective lagrangians to describe the LHC data?
• new state belongs to SU(2) doublet
• consider SU(2) x U(1) invariant dimension-6 lagrangian

Leff = LSM + X

n

fn Λ2 On + · · ·

• There are 59 “independent” dimension-six operators
• our choice for the boson operators is

[Buchmuller & Wyler; Grzadkowski] [Corbett, OE, Gonzalez-Fraile, Gonzalez-Garcia]

OGG = Φ†Φ Ga

µνGaµν

OW W = Φ† ˆ Wµν ˆ W µνΦ OBB = Φ† ˆ Bµν ˆ BµνΦ OBW = Φ† ˆ Bµν ˆ W µνΦ OW = (DµΦ)† ˆ W µν(DνΦ) OB = (DµΦ)† ˆ Bµν(DνΦ) OΦ,1 = (DµΦ)† Φ Φ† (DµΦ) OΦ,2 = 1 2∂µ Φ†Φ

• ∂µ
• Φ†Φ
• OΦ,4 = (DµΦ)† (DµΦ)
• Φ†Φ
• constrained by EWPD

DµΦ =

• ∂µ + ig0Bµ/2 + igσaW a

µ/2

• Φ

ˆ Bµν = ig0Bµν/2 ˆ Wµν = igσaW a

µν/2

with

LHV V

eff

= − αs 8π fGG Λ2 OGG + fBB Λ2 OBB + fW W Λ2 OW W + fB Λ2 OB + fW Λ2 OW + fΦ,2 Λ2 OΦ,2

eliminated with EOM

[Hagiwara, Ishihara, Szalapski, Zeppenfeld]

LHV V = gHgg HGa

µνGaµν + gHγγ HAµνAµν + g(1) HZγ AµνZµ∂νH + g(2) HZγ HAµνZµν

+ g(1)

HZZ ZµνZµ∂νH + g(2) HZZ HZµνZµν + g(3) HZZ HZµZµ

+ g(1)

HW W

• W +

µνW − µ∂νH + h.c.

• + g(2)

HW W HW + µνW − µν + g(3) HW W HW + µ W − µ

• 10 effective HVV couplings are generated

that depend on 6 Wilson coefficients

gHgg = −αs 8π fGGv Λ2 g(1)

HZγ = g2v

2Λ2 sw(fW − fB) 2cw gHγγ = −g2vs2

w

2Λ2 fBB + fW W 2 g(2)

HZγ = g2v

2Λ2 sw(2s2

wfBB − 2c2 wfW W )

2cw g(1)

HZZ = g2v

2Λ2 c2

wfW + s2 wfB

2c2

w

g(1)

HW W = g2v

2Λ2 fW 2 g(2)

HZZ = − g2v

2Λ2 s4

wfBB + c4 wfW W

2c2

w

g(2)

HW W = − g2v

2Λ2 fW W g(3)

HZZ = m2 Z(

√ 2GF )1/2 ✓ 1 − v2 2Λ2 fΦ,2 ◆ g(3)

HW W = m2 W (

√ 2GF )1/2 ✓ 1 − v2 2Λ2 fΦ,2 ◆

Hgg Hγγ HγZ HZZ HW +W − γW +W − ZW +W − OGG √ OBB √ √ √ OW W √ √ √ √ OB √ √ √ √ OW √ √ √ √ √ OΦ,2 √ √

• Summarizing

coefficients related by gauge invariance potential correlations

• we should also include fermionic operators for the third generation

OeΦ,33 = (Φ†Φ)(¯ L3ΦeR,3) OuΦ,33 = (Φ†Φ)( ¯ Q3 ˜ ΦuR,3) OdΦ,33 = (Φ†Φ)( ¯ Q3ΦdR,3)

leading to

LHff

eff

=fτmτ vΛ2 OeΦ,33 + fbmb vΛ2 OdΦ,33 + ftmt vΛ2 OuΦ,33 + fΦ,2 Λ2 OΦ,2

LHff = gfH ¯ fLfR + h.c. with gf = −mf v ✓ 1 − v2 2Λ2 fΦ,2 − v2 √ 2Λ2 ff ◆

implying that

Rate-based analysis

• decays and cross sections evaluated with FeynRules+MadGraph
• SM K-factors
• difference to previous analyses:
• there are new correlations in addition to fGG × ft

ft

fWW/Λ2 [TeV-2] fBB/Λ2 [TeV-2]

• 10
• 5

5 10 15 20

• 20
• 15
• 10
• 5

5 10

• 10
• 5

5 10 15 20

• 20
• 15
• 10
• 5

5 10

fW/Λ2 [TeV-2] fB/Λ2 [TeV-2]

• 20
• 10

10 20 30

• 80
• 60
• 40
• 20

20 40

• 20
• 10

10 20 30

• 80
• 60
• 40
• 20

20 40

strongest correlation due to H → γγ suppressed by improved with

fB s2

w

H → Zγ

∆(-2 ln L)

2 4 6 8 10 12 14 16

• the 9 parameter fit leads to
• 60
• 50
• 40
• 30
• 20
• 10

10 20

OGG OWW OBB OW OB Oφ2

0.15 0.2 0.25 0.3 0.5 ∞ 0.5 0.3 0.25 f/Λ2 [TeV-2] Λ/ √|f| [TeV]

L=4.5-5.1(7 TeV)+19.4-20.3(8 TeV) fb-1, ATLAS + CMS

SM exp. 68% CL SM exp. 95% CL data 68% CL data 95% CL

• 20

20 40 60

Ot Ob Oτ

0.2 0.25 0.3 0.5 ∞ 0.5 0.3 0.25 0.2 0.15 f/Λ2 [TeV-2] Λ/ √|f| [TeV]

L=4.5-5.1(7 TeV)+19.4-20.3(8 TeV) fb-1, ATLAS + CMS

• strongest constraints on
• next are

fW W and fBB fW and fφ,2

Kinematic distributions

• To avoid double counting we used asymmetries in VH:

Ai = bini+1 − bini bini+1 + bini

we adopted the 1st and 3rd strategies

• lead to new Lorentz structures
• limited to fully documented distributions
• using VH and WBF
• unitarity might be an issue:

take the results at face value (with care!) introduce ad-hoc form factors keep only the phase space region not sensitive to UV completion

(OB , OW , OBB , OW W )

• FeynRules + MadGraph + Pythia + PGS4/DELPHES

[Ellis & Sanz & You;…],

• we used ATLAS results for VH (0,1,2 leptons) and WBF γγjj

pTV(GeV) Events/bin SM (SM Higgs) x 70 (fW/Λ2 =20TeV-2) x 70 2leptons 50 100 150 200 250 103 102

∆Φjj Events/bin SM Higgs fWW/Λ2=-fBB/Λ2=20TeV-2 fWW/Λ2=-fBB/Λ2=-20TeV-2 γγ jj π/3 2π/3 π 10 102

• for the WBF analysis

A1 =σ(∆φjj < π

3 ) + σ(∆φjj > 2π 3 ) − σ( π 3 < ∆φjj < 2π 3 )

σ(∆φjj < π

3 ) + σ(∆φjj > 2π 3 ) + σ( π 3 < ∆φjj < 2π 3 ) ,

A2 =σ(∆φjj > 2π

3 ) − σ(∆φjj < π 3 )

σ(∆φjj > 2π

3 ) + σ(∆φjj < π 3 ) ,

A3 =σ(∆φjj > 5π

6 ) − σ( 2π 3 < ∆φjj < 5π 6 )

σ(∆φjj > 5π

6 ) + σ( 2π 3 < ∆φjj < 5π 6 )

fWW/Λ2 [TeV-2] fBB/Λ2 [TeV-2]

• 10
• 5

5 10 15 20

• 20
• 15
• 10
• 5

5 10

• 10
• 5

5 10 15 20

• 20
• 15
• 10
• 5

5 10 fW/Λ2 [TeV-2] fB/Λ2 [TeV-2]

• 20
• 10

10 20 30

• 80
• 60
• 40
• 20

20 40

• 20
• 10

10 20 30

• 80
• 60
• 40
• 20

20 40 fWW/Λ2 [TeV-2] fBB/Λ2 [TeV-2]

• 10
• 5

5 10 15 20

• 20
• 15
• 10
• 5

5 10

• 10
• 5

5 10 15 20

• 20
• 15
• 10
• 5

5 10 fW/Λ2 [TeV-2] fB/Λ2 [TeV-2]

• 20
• 10

10 20 30

• 80
• 60
• 40
• 20

20 40

• 20
• 10

10 20 30

• 80
• 60
• 40
• 20

20 40 fWW/Λ2 [TeV-2] fBB/Λ2 [TeV-2]

• 10
• 5

5 10 15 20

• 20
• 15
• 10
• 5

5 10

• 10
• 5

5 10 15 20

• 20
• 15
• 10
• 5

5 10 fW/Λ2 [TeV-2] fB/Λ2 [TeV-2]

• 20
• 10

10 20 30

• 80
• 60
• 40
• 20

20 40

• 20
• 10

10 20 30

• 80
• 60
• 40
• 20

20 40

WBF all

• last bin

• Presently the distributions impact mainly and

OB OW

• 50
• 40
• 30
• 20
• 10

O

G G

O

W W

O

B B

O

W

O

B

O

φ 2

0.15 0.2 0.25 0.3 0.5 ∞ 0.5 f/Λ2 [TeV-2] Λ/ √|f| [TeV]

L=4.5-5.1(7 TeV)+19.4-20.3(8 TeV) fb-1, 68% CL: ATLAS + CMS

rate only rate+distributions

• 15
• 10
• 5

5

O

t

O

b

O

τ

0.3 0.4 0.5 1 ∞ 1 0.5 f/Λ2 [TeV-2] Λ/ √|f| [TeV]

L=4.5-5.1(7 TeV)+19.4-20.3(8 TeV) fb-1, 68% CL: ATLAS + CMS

take this is a proof of principle for the use of distributions!

• EFT range of validity from unitarity violation [Corbett, OE, Gonzalez-Garcia]
• fΦ2

Λ2 s

• ≤

105 = ⇒ √s < 2.3 TeV

• fW

Λ2 s

• ≤

205 = ⇒ √s < 5.3 TeV

• fB

Λ2 s

• ≤

640 = ⇒ √s < 3.7 TeV

• fW W

Λ2 s

• ≤

200 = ⇒ √s < 2.6 TeV

• fBB

Λ2 s

• ≤

880 = ⇒ √s < 9.4 TeV

V V → V V and f ¯ f → V V

[Renard & Gounaris; Baur & Zeppenfeld; ….]

OW OB

• the operators and modify TGC's
• 3. TGC from Higgs measurements

[Corbett, OE, Gonzalez-Fraile, Gonzalez-Garcia: arXiv:1304.1151]

∆κγ = g2v2 8Λ2 ⇣ fW + fB ⌘ ∆gZ

1

= g2v2 8c2Λ2 fW ∆κZ = g2v2 8c2Λ2 ⇣ c2fW − s2fB ⌘

OW OB

• the operators and modify TGC's
• 3. TGC from Higgs measurements

[Corbett, OE, Gonzalez-Fraile, Gonzalez-Garcia: arXiv:1304.1151]

∆κγ = g2v2 8Λ2 ⇣ fW + fB ⌘ ∆gZ

1

= g2v2 8c2Λ2 fW ∆κZ = g2v2 8c2Λ2 ⇣ c2fW − s2fB ⌘

• nice interplay between TGC and Higgs physics
• side effect: TGC data still not good enough to decouple and OW

OB

rate analysis rate+distributions

OW OB

• the operators and modify TGC's
• 3. TGC from Higgs measurements

[Corbett, OE, Gonzalez-Fraile, Gonzalez-Garcia: arXiv:1304.1151]

∆κγ = g2v2 8Λ2 ⇣ fW + fB ⌘ ∆gZ

1

= g2v2 8c2Λ2 fW ∆κZ = g2v2 8c2Λ2 ⇣ c2fW − s2fB ⌘

• nice interplay between TGC and Higgs physics
• side effect: TGC data still not good enough to decouple and OW

OB

rate analysis rate+distributions

• 4. Off-shell Higgs measurements
• Off-shell Higgs measurements are a window into its width:

σon-shell

i→H→f ∝

g2

i (mH) g2 f(mH)

ΓH vs σoff-shell

i→H∗→f ∝ g2 i (m4`) g2 f(m4`)

[Kauer & Passarino; Caola & Melnikov; Ellis & Williams]

• is also useful to break the degeneracy

m4`

∆t × ∆g

[Buschmann, Gonçalves, Kuttimalai, Schönherr, Kraus, Plehn]

L = LSM + ∆W gmW H W µWµ + ∆Z g 2cw mZH ZµZµ − X

τ,b,t

∆f mf v H ¯ fRfL + h.c.

• + ∆gFG

H v GµνGµν + ∆γFA H v AµνAµν + invisible decays + unobservable decays

• In this analysis we worked in the delta framework

additional contribution!

• sample of Feynman diagrams

leading to

Mgg→ZZ = (1 + ∆Z) [(1 + ∆t)Mt + ∆gMg] + Mc dσ dm4` = (1 + ∆Z)  (1 + ∆t) dσtc dm4` + ∆g dσgc dm4`

• + (1 + ∆Z)2

 (1 + ∆t)2 dσtt dm4` + (1 + ∆t)∆g dσtg dm4` + ∆2

g

dσgg dm4`

• + dσc

dm4`

sensitive to the sign!

• top mass effects are different for ∆g and ∆t

• sample of Feynman diagrams

leading to

Mgg→ZZ = (1 + ∆Z) [(1 + ∆t)Mt + ∆gMg] + Mc dσ dm4` = (1 + ∆Z)  (1 + ∆t) dσtc dm4` + ∆g dσgc dm4`

• + (1 + ∆Z)2

 (1 + ∆t)2 dσtt dm4` + (1 + ∆t)∆g dσtg dm4` + ∆2

g

dσgg dm4`

• + dσc

dm4`

sensitive to the sign!

• top mass effects are different for ∆g and ∆t

∆g ∆t

• 4
• 3
• 2
• 1

1 2 3 4

• 4
• 3
• 2
• 1

1 2 ∆g ∆t

• 4
• 3
• 2
• 1

1 2 3 4

• 4
• 3
• 2
• 1

1 2 ∆g ∆t

• 4
• 3
• 2
• 1

1 2 3 4

• 4
• 3
• 2
• 1

1 2 ∆g ∆t

• 4
• 3
• 2
• 1

1 2 3 4

• 4
• 3
• 2
• 1

1 2

without off-shell data with off-shell data

• it is interesting to look at 1d profile likelihoods
• 3
• 2
• 1

1 2 4 6 8 10 12

On+Off-Shell On-Shell

t

∆ (-2 ln L) ∆ 10 20 30 2 4 6 8 10

SM H

Γ /

H

Γ (-2 ln L) ∆

SM+dim6

68% CL 95% CL Run I LHC - ATLAS+CMS

slight preference for flipped-sign solution ΓH < 9.3 ΓSM

H

at 68%CL

• 5. Final remarks
• The observed Higgs is fully consistent with the SM
• kinematic distributions and off-shell measurements can play a major

role

• We tested different ways to include the theoretical uncertainties.

BRinv < 0.31 at 95% CL ΓH < 9.3 ΓSM

H

at 68%CL

• 5. Final remarks
• The observed Higgs is fully consistent with the SM
• kinematic distributions and off-shell measurements can play a major

role

• We tested different ways to include the theoretical uncertainties.

BRinv < 0.31 at 95% CL ΓH < 9.3 ΓSM

H

at 68%CL

THANK YOU

backup slides

• We can also analyze distributions in the non-linear sigma model

[Brivio, et al, arXiv: 1311.1823]

• 1
• 0.5

0.5 1

aG*102 aB aW a4 a5 cH a’t a’b a’τ L=4.5-5.1(7 TeV)+19.4-20.3(8 TeV) fb-1, ATLAS + CMS

Rates 68% CL Rates 95% CL Rates+dist. 68% CL Rates+dist. 95% CL

There are many operators at this order

generalization of the Appelquist-Bernard-Longhitano basis TGV T S QGV QGV QGV

• Limits on the Higgs couplings

Rate-based analysis

fWW/Λ2 [TeV-2] fBB/Λ2 [TeV-2]

• 10
• 5

5 10 15 20

• 20
• 15
• 10
• 5

5 10

• 10
• 5

5 10 15 20

• 20
• 15
• 10
• 5

5 10

fB/Λ2 [TeV-2] fBB/Λ2 [TeV-2]

• 80
• 60
• 40
• 20

20 40

• 20
• 15
• 10
• 5

5 10

• 80
• 60
• 40
• 20

20 40

• 20
• 15
• 10
• 5

5 10 fW/Λ2 [TeV-2] fB/Λ2 [TeV-2]

• 20
• 10

10 20 30

• 80
• 60
• 40
• 20

20 40

• 20
• 10

10 20 30

• 80
• 60
• 40
• 20

20 40

strongest correlation due to H → γγ suppressed by improved with

fB s2

w

H → Zγ

• decays and cross sections evaluated with FeynRules+MadGraph
• SM K-factors
• difference to previous analyses:
• there are new correlations in addition to fGG × ft

ft