[PPT] - Higgs coupling measurements with ATLAS Richard Mudd University of PowerPoint Presentation

SLIDE 1

Higgs coupling measurements with ATLAS

Richard Mudd

University of Birmingham

HEP Seminar, Birmingham 12th November 2014

SLIDE 2

July 2012

2 of 39

SLIDE 3

Higgs Mechanism

SU(2)L ⊗ U(1)Y describes

electroweak sector in terms of massless gauge bosons

In the SM a complex scalar doublet is

introduced φ = φ+ φ0

For Higgs mechanism potential

chosen such that electroweak symmetry is hidden

Higgs field gets non-zero vacuum

expectation value

Three degrees of freedom give

W +, W −, Z mass, one gives new scalar boson - the Higgs boson

Image credit: Philip Tanedo

3 of 39

SLIDE 4

Higgs Mechanism: Scalar Couplings Structure

Bosonic sector:

EWSB gives mass to W +, W −, Z bosons
Higgs couplings proportional to m2

W /Z

gHVV = 2m2

V

v

H V V gHV V

gHf ¯

f

H f ¯ f

Fermionic sector:

After introducting Higgs field, can add

Yukawa terms to Lagrangian

Higgs couplings proportional to fermion mass

gHf ¯

f = Yf = mf

v

v is Higgs field vacuum expectation value
Loops (e.g. γ, gluon) sensitive to BSM physics

4 of 39

SLIDE 5

Higgs Production at the LHC

[GeV]

H

M 80 100 200 300 400 500 1000 H+X) [pb] → (pp σ

2

10

1

10 1 10 = 8 TeV s

LHC HIGGS XS WG 2014

H (NNLO+NNLL QCD + NLO EW) → pp qqH (NNLO QCD + NLO EW) → pp W H ( N N L O Q C D + N L O E W ) → p p

Z H ( N N L O Q C D + N L O E W ) → p p

t t H ( N L O Q C D ) → p p

bbH (NNLO QCD in 5FS, NLO QCD in 4FS) → pp

Gluon fusion mode dominates
Subleading modes essential to tag more difficult

decay modes and measure couplings

g g ¯ t t t H q q W/Z W/Z H q ¯ q W ∗/Z∗ W/Z H

H ¯ t t ¯ t t g g 5 of 39

SLIDE 6

Higgs Decays at the LHC

H → b¯

b has highest rate but challenging due to very large background

H → WW (∗) → lνlν,

H → ττ also have relatively high rates but complex final states

H → ZZ (∗) → 4ℓ,

H → γγ challenging because of low rates but clean final states [GeV]

H

M

80 100 120 140 160 180 200

Higgs BR + Total Uncert

4

10

3

10

2

10

1

10 1

LHC HIGGS XS WG 2013

b b τ τ µ µ c c gg γ γ γ Z WW ZZ

6 of 39

SLIDE 7

Possible Extensions to SM Higgs Sector

In the SM EWSB is achieved through a single complex scalar doublet but

many extensions possible Additional EW singlet

Mixing between singlet original Higgs doublet → two CP-even bosons
Couple to SM particles in a similar way to SM Higgs

Two Higgs Doublet

Predict 5 Higgs Bosons: 2 neutral CP-even, one neutral CP odd, 2 charged
e.g. MSSM
Typically require that models satisfy Glashow-Weinberg condition, e.g:
Type I: one doublet couples to vector bosons, one to fermions
Type II: one doublet couples to up-type quarks, the other to down-type and leptons

7 of 39

SLIDE 8

How does new physics modify Higgs couplings?

New physics (e.g. extended Higgs sectors) can modify the Higgs couplings
Modifications depend on mass scale of new physics
For new physics at 1 TeV scale modifications are typically ∼ 1 - 10 %

Model κV κb κγ Singlet mixing ∼ 6% ∼ 6% ∼ 6% 2HDM ∼ 1% ∼ 10% ∼ 1% Decoupling MSSM ∼ -0.001% ∼ 1.6% ∼ -0.4% Composite ∼ -3% ∼ -(3-9)% ∼ -9% Top Partner ∼ -2% ∼ -2% ∼ +1% From Snowmass Higgs Working Group Report

8 of 39

SLIDE 9

ATLAS detector

Successful operation of ATLAS

detector in run I

4.6 fb−1 at √s = 7TeV ,

20.3 fb−1 at √s = 8TeV

≃ 95% of recorded luminosity

good for physics

Strong detector

performance achieved in challenging environment

Average 21

interactions per bunch crossing

Higher than design

pileup

Month in Year J a n A p r J u l O c t ]

1

Delivered Luminosity [fb 5 10 15 20 25 30 35

= 7 TeV s 2010 pp = 7 TeV s 2011 pp = 8 TeV s 2012 pp

ATLAS Online Luminosity Mean Number of Interactions per Crossing 5 10 15 20 25 30 35 40 45 /0.1]

1

Recorded Luminosity [pb 20 40 60 80 100 120 140 160 180 Online Luminosity ATLAS

> = 20.7 µ , <

1

Ldt = 21.7 fb

∫

= 8 TeV, s > = 9.1 µ , <

1

Ldt = 5.2 fb

∫

= 7 TeV, s

9 of 39

SLIDE 10

Atlas Higgs physics programme

ATLAS has published a broad selection of results in the Higgs sector in run I
Mass
Couplings
Spin/CP
Differential distributions
Rare decays
and more ...
Focus on measurement of coupling properties today
Don’t have time to discuss individual analyses in detail
Instead a selection of highlights from main inputs to ATLAS combined coupling

measurements

For bb see Paul Thompson’s recent seminar

10 of 39

SLIDE 11

ATLAS Higgs couplings measurements

ATLAS has recently released updated results for the five most sensitive SM channels using full run I data:

H → 4ℓ
H → γγ
VH, H → b¯

b

H → WW
H → ττ

Signal strength

1

1 2 3 4 5 6 7 8 ATLAS = 7 TeV s ,

1

Ldt = 4.5 fb

∫

= 8 TeV s ,

1

Ldt = 20.3 fb

∫

= 125.4 GeV

H

m , γ γ → H Total Stat. Syst. µ

ggF

µ

VBF

µ

WH

µ

ZH

µ

H t t

µ

[GeV]

H

m 110 115 120 125 130 135 140

SM

σ / σ 95% C.L. limit on 0.5 1 1.5 2 2.5 3 3.5 4 Observed (CLs) Expected (no Higgs) = 125 GeV)

H

Expected (m σ 1 ± σ 2 ±

ATLAS

1

Ldt = 20.3 fb

∫

= 8 TeV, s

1

Ldt = 4.7 fb

∫

= 7 TeV, s

ggF

µ 0.5 1 1.5 2 2.5

VBF

µ 0.5 1 1.5 2 2.5 3 3.5 4 Λ

2 ln

2 4 6 8 10 12 14

σ 1 σ 2 σ 3

(1.00,1.27)

SM

Preliminary ATLAS ν l ν l → WW* → H

1

Ldt = 20.3 fb

∫

= 8 TeV s

1

Ldt = 4.5 fb

∫

= 7 TeV s Best Fit SM

[GeV]

H

m 120 122 124 126 128 130 Local p

Obs 2012 Exp 2012 Obs 2011 Exp 2011 Obs combination Exp combination ATLAS l 4 → ZZ* → H

1

Ldt = 4.5 fb

∫

=7 TeV s

1

Ldt = 20.3 fb

∫

=8 TeV s σ 2 σ 4 σ 6 σ 8

15

10

12

10

9

10

6

10

3

10 1

(S / B)

10

log

4
3
2
1

1 Events / bin 1 10

2

10

3

10

4

10

Preliminary ATLAS

1

, 20.3 fb = 8 TeV s

1

, 4.5 fb = 7 TeV s τ τ → H

Data =1.4) µ Background ( =0) µ Background ( =1.4) µ ( τ τ → (125) H =1) µ ( τ τ → (125) H

11 of 39

SLIDE 12

‘Signal Strength’ µ

Measured rates reported relative to SM prediction
Signal strength defined as:

µ = σ · BR σSM · BRSM

Measured in decay modes and also for their combination
Also able to measure rates for specific production modes
Typically denoted with a subscript

µggF = σ(ggF) · BR σSM(ggF) · BRSM

Often combine bosonic/fermionic production modes
µggF+ttH, µVBF+VH

12 of 39

SLIDE 13

Statistical techniques

Confidence intervals based on profile likelihood ratio

Λ(α) = L

α, ˆ

ˆ θ(α)

L(ˆ

α, ˆ θ) = Maximum likelihood for given α Global maximum likelihood

Depends on one of more parameters of interest, α
e.g. (µ, mH), (µggF ,µVBF )
Systematic uncertainties modelled using nuisance parameters, θ
Typically constrained by gaussians
Model uncertainties and their correlations
Likelihood functions built using sums of signal and background pdfs in

discriminating variables

13 of 39

SLIDE 14

H → ZZ (∗) → 4ℓ analysis

Low rates but final state with good mass resolution (1.6 - 2.2 GeV) and high

S/B (0.7 - 1.8)

σ × BR ≃ 2.9 fb for mH = 125.5 GeV
Two same-flavour, opposite sign lepton pairs
Low pT electron/muon performance critical
pT > 7 (6) GeV for electrons (muons)
Isolation and impact parameter requirements to

reduce background

[GeV]

µ 2e/2e2 µ 2

m 80 100 120 140 / 0.5 GeV

µ 2e/2e2 µ 2

1/N dN/dm 0.02 0.04 0.06 0.08 0.1

= 125 GeV

H

m Gaussian fit

Simulation ATLAS

µ 2e/2e2 µ 2 → ZZ* → H = 8 TeV s 0.01 GeV ± m = 124.78 0.01 GeV ± = 1.77 σ : 20% σ 2 ± Fraction outside With Z mass constraint

mZ constrained

kinematic fit for m12

FSR photon recovery

for m12 candidates

E-p combination for

pe

T < 30 GeV

[GeV]

T

E 10 20 30 40 50 60 70 80 90 100 Efficiency 0.6 0.65 0.7 0.75 0.8 0.85 0.9 0.95 1

1

L dt = 20.3 fb

∫

ee → = 8 TeV Z s ATLAS Preliminary 2012 | < 2.47 η | LooseLLH LooseLLH, MC VeryTightLLH VeryTightLLH, MC

5 10 15 20 25 30 35 40 45 50

Efficiency 0.9 0.92 0.94 0.96 0.98 1 ATLAS Preliminary

1

Ldt =2264 pb

∫

2012 data, chain 3 >=17.3 µ < MC data µ 5 10 15 20 25 30 35 40 45 50 Data/MC 0.98 0.99 1 1.01 1.02

14 of 39

SLIDE 15

H → ZZ (∗) → 4ℓ categorisation and fit model

Multi-observable fit in production-tagged categories
Exploit use of BDTs

ATLAS

l 4 → ZZ* → H

selection l 4 High mass two jets VBF VBF enriched Low mass two jets jj)H → jj)H, Z( → W( Additional lepton )H ll → )H, Z( ν l → W( VH enriched ggF ggF enriched

ggF categories:

Fit m4ℓ and BDT with LO matrix element

kinematic discriminant, p4ℓ

T , η4ℓ

VBF category:

Fit m4ℓ and BDT with jet kinematic variables

VH categories:

1D fit to m4ℓ
utput

ZZ*

D

6
4
2

2 4 6 8

utput / 0.5

ZZ*

1/N dN/dD 0.05 0.1 0.15 0.2 0.25

ATLAS Simulation l 4 → ZZ* → H

1

Ldt = 4.5 fb

∫

= 7 TeV s

1

Ldt = 20.3 fb

∫

= 8 TeV s =125 GeV)

H

ggF (m ZZ*

[GeV]

jj

m 200 400 600 800 1000 / 10 GeV

jj

1/N dN/dm 0.02 0.04 0.06 0.08 0.1

Simulation

ATLAS

l 4 → ZZ* → H

1

Ldt = 4.5 fb

∫

= 7 TeV s

1

Ldt = 20.3 fb

∫

= 8 TeV s category VBF enriched =125 GeV

H

m

ggF VBF

15 of 39

SLIDE 16

H → ZZ (∗) → 4ℓ categorisation and fit model

utput

VBF

BDT

1 -0.8-0.6-0.4-0.2 0 0.2 0.4 0.6 0.8 1

Purity (S/(S+B)) 0.2 0.4 0.6 0.8 1 1.2 ATLAS

l 4 → ZZ* → H

1

Ldt = 4.5 fb

∫

= 7 TeV s

1

Ldt = 20.3 fb

∫

= 8 TeV s category VBF enriched [GeV] < 140

l 4

110 < m =125 GeV

H

m

H, VH Purity t H, t b ggF, VBF, b VBF purity

VBF+VH

µ 0.5 1 1.5 2 2.5 3 3.5 4 4.5 Λ

2ln

0.5 1 1.5 2 2.5 3 3.5 4

ATLAS l 4 → ZZ* → H

1

Ldt = 4.5 fb

∫

= 7 TeV s

1

Ldt = 20.3 fb

∫

= 8 TeV s

ATLAS-CONF-2013-013 ggF and VBF enriched categories ggF, VBF and VH-Hadronic enriched categories All categories SM

B/B ×

H t H+t b ggF+b

µ 0.5 1 1.5 2 2.5 3 3.5 4

SM

B/B ×

VH+VBF

µ

4
2

2 4 6 8 10 12 14

SM Best Fit 68% CL 95% CL ATLAS l 4 → ZZ* → H

1

Ldt =4.5 fb

∫

=7 TeV s

1

Ldt =20.3 fb

∫

=8 TeV s 2D model ggF = 125.36 GeV

H

m

5 events in VBF-enriched category, 1 with BDTVBF ≃ 0.7

µggF+bbH+ttH × B/BSM = 1.7+0.5

−0.4

µVBF+VH × B/BSM = 0.3+1.6

−0.9

Uncertainties dominated by statistical component
Expected uncertainty on µVBF+VH reduced by ≃ 40% compared to preliminary

result

16 of 39

SLIDE 17

H → γγ analysis

H → γγ decays through t and W loops in SM
Negative interference between t and W

contributions

Two isolated, high pT photons
Search for narrow peak (mass resolution 1.3 -

1.8 GeV) on top of background (S/B ≃ 3%)

[GeV]

γ γ

m 110 115 120 125 130 135 140 / 0.5 GeV

γ γ

/dm 1/N dN 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16

ATLAS Simulation = 8 TeV s = 125 GeV

H

, m γ γ → H

Tt

p Central - high

Tt

p Forward - low MC Model MC Model (tight)

ID

ε 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 | < 0.6 η | ≤ γ unconverted < 4 GeV

iso T

E = 8 TeV s

1

Ldt = 20.3 fb

∫

ATLAS Preliminary Electron extrapolation Matrix method γ ll → Z [GeV]

T

E 20 30 40

2

10

2

10 × 2

ID

ε > -

ID

ε <

0.1

0.1 (tight)

ID

ε 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 | < 1.37 η | ≤ 0.6 γ unconverted < 4 GeV

iso T

E = 8 TeV s

1

Ldt = 20.3 fb

∫

ATLAS Preliminary Electron extrapolation Matrix method γ ll → Z [GeV]

T

E 20 30 40

2

10

2

10 × 2

ID

ε > -

ID

ε <

0.1

0.1 (tight)

ID

ε 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 | < 1.81 η 1.52 < | γ unconverted < 4 GeV

iso T

E = 8 TeV s

1

Ldt = 20.3 fb

∫

ATLAS Preliminary Electron extrapolation Matrix method γ ll → Z [GeV]

T

E 20 30 40

2

10

2

10 × 2

ID

ε > -

ID

ε <

0.1

0.1 (tight)

ID

ε 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 | < 2.37 η | ≤ 1.81 γ unconverted < 4 GeV

iso T

E = 8 TeV s

1

Ldt = 20.3 fb

∫

ATLAS Preliminary Electron extrapolation Matrix method γ ll → Z [GeV]

T

E 20 30 40

2

10

2

10 × 2

ID

ε > -

ID

ε <

0.1

0.1

Diphoton invariant mass:

m2

γγ = 2E1E2(1 − cosα)

Neural network based

identification of primary interaction vertex

Backgrounds γγ(75%), γj, jj
Estimated from sideband fit

17 of 39

SLIDE 18

H → γγ categories

Comprehensive categorisation scheme targetting 5 main production mechanisms

Signal strength

2

2 4 6 8 10

ATLAS = 7 TeV s ,

1

Ldt = 4.5 fb

∫

= 8 TeV s ,

1

Ldt = 20.3 fb

∫

= 125.4 GeV

H

m , γ γ → H

Combined leptonic H t t hadronic H t t

ne-lepton

VH

T miss

E VH di-jet VH VBF tight VBF loose

Tt

p Forward high

Tt

p Forward low

Tt

p Central high

Tt

p Central low

Untagged:

Split based on pTt and position in detector

VBF:

Cut on output of BDT
Loose and tight categories

VH:

Sensitivity to separate WH and ZH
Hadronic, leptonic and E miss

T

signatures ttH:

Hadronic and leptonic top decays

Signal strength for each production mode consistent with SM

Signal strength

1

1 2 3 4 5 6 7 8 ATLAS = 7 TeV s ,

1

Ldt = 4.5 fb

∫

= 8 TeV s ,

1

Ldt = 20.3 fb

∫

= 125.4 GeV

H

m , γ γ → H Total Stat. Syst. µ

ggF

µ

VBF

µ

WH

µ

ZH

µ

H t t

µ

18 of 39

SLIDE 19

H → WW (∗) → ℓνℓν analysis

High rate, relatively clean final state (ee, eµ, µµ with E miss

T

/pmiss

T

)

Mass resolution ≃ 15 GeV

⇒ background control crucial

Several background sources
WW,W+jets,tt,single top, Zγ∗, Z→ ℓℓ estimated

in data using control regions

Other diboson process estimate using MC
Background composition depends on lepton

flavour, Njets

Improvements with respect to preliminary analysis
Track-based missing ET
Electron Likelihood ID
Reduce lepton ET threshold 15 → 10 GeV
Optimised event categorisation
Overall 30% reduction of uncertainties on µ

w.r.t preliminary results

Unit normalized 0.01 0.02 0.03

100
50

50 100 Unit normalized 0.02 0.04 0.06

miss T

p RMS=12.4

miss T

E RMS=15.9

miss T

p Using RMS=14.1

miss T

E Using RMS=18.8

(b) [GeV]

T

m

Reco. - Gen. for

(a) [GeV]

miss T

E

r

miss T

p

Reco. - Gen. for

ATLAS Simulation Prelim.

WW* → MC sample for ggF H 19 of 39

SLIDE 20

H → WW (∗) → ℓνℓν categories and fit model

Transverse mass mT used as discriminant in fit
In VBF categories use BDT instead
Fit in several signal and control regions
Rates for ggF and VBF processes consistent with SM
Observe VBF production with 3.2σ significance

nj = 0 nj = 1 nj ≥ 2 enriched VBF- ggF- enriched ee/µµ ee/µµ eµ VBF-enriched selection Pre- eµ eµ (8 TeV) ee/µµ eµ ggF-enriched ggF

µ 0.5 1 1.5 2 2.5

VBF

µ 0.5 1 1.5 2 2.5 3 3.5 4 Λ

2 ln

2 4 6 8 10 12 14

σ 1 σ 2 σ 3

(1.00,1.27)

SM

Preliminary ATLAS ν l ν l → WW* → H

1

Ldt = 20.3 fb

∫

= 8 TeV s

1

Ldt = 4.5 fb

∫

= 7 TeV s Best Fit SM

ggF

µ /

VBF

µ 1 2 3 4 5 Λ

2 ln

2 4 6 8 10 σ 1 σ 2 σ 3

(1.25,0.0) (1.25,0.0) (0.73,1.00) (2.04,1.00) (0.36,4.00) (3.37,4.00)

Preliminary ATLAS ν l ν l → WW* → H

1

Ldt = 20.3 fb

∫

= 8 TeV s

1

Ldt = 4.5 fb

∫

= 7 TeV s

20 of 39

SLIDE 21

H → ττ analysis

Three final states used in analysis

depending on τ decays:

τlepτlep
τlepτhad
τhadτhad
Z → ττ and fake τ backgronds dominate
Use missing mass calculator
Use visible τ decay products and E miss

T

to find most-likely mττ

Z → ττ background from Z → µµ

embedding method

BDT used as a discriminating variable in

a 6 category (VBF and boosted for each final state) fit

Cut-based analysis as cross check

Vtx

N

5 10 15 20 25 30

Signal Efficiency

0.2 0.4 0.6 0.8 1 1.2

TauBDT loose TauBDT medium TauBDT tight Multi Prong | < 2.5 η > 15 GeV, |

T

p

= 8 TeV s Simulation 2012, Preliminary ATLAS

MMC [GeV]

60 80 100 120 140 160 180 Arbitrary Units 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 MC MC Stat. Error Embedded MC

Emb. Uncertainty

ATLAS Simulation Preliminary

[GeV]

MMC τ τ

m

60 80 100 120 140 160 180

Emb. / MC

0.8 0.9 1 1.1 1.2 21 of 39

SLIDE 22

H → ττ: evidence for Higgs decays to fermions

Direct evidence for

coupling to fermions at 4.5σ level (3.5σ exp)

µ = 1.42+0.44

−0.38 consistent

with SM Yukawa coupling prediction

ln(1+S/B) w. Events / 10 GeV

20 40 60 80

[GeV]

τ τ MMC

m 50 100 150 200

weighted (Data-Bkg.)

10 20

=1.4) µ ( (125) H =1.6) µ ( (110) H =6.2) µ ( (150) H Data =1.4) µ ( (125) H τ τ → Z Others Fakes Uncert.

ATLAS Preliminary VBF+Boosted τ τ → H

1

, 4.5 fb = 7 TeV s

1

, 20.3 fb = 8 TeV s

(S / B)

10

log

4
3
2
1

1 Events / bin 1 10

2

10

3

10

4

10

Preliminary ATLAS

1

, 20.3 fb = 8 TeV s

1

, 4.5 fb = 7 TeV s τ τ → H

Data =1.4) µ Background ( =0) µ Background ( =1.4) µ ( τ τ → (125) H =1) µ ( τ τ → (125) H

[GeV]

µ

+

µ

m 110 115 120 125 130 135 140 145 150 155 160 Events / 2.5 GeV 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000

ATLAS

1

Ldt = 24.8 fb

∫

= 8 TeV s = 7 TeV s

µ

+

µ → H

Data MC (stat)

*

γ Z/ γ WZ/ZZ/W t t WW Single Top W+jet 300 × H [125 GeV]

ATLAS also searches for H → µµ
No observed excess of events
In SM BR(ττ)/BR(µµ) ≃ 300

⇒ The Higgs does not couple universally to different flavour leptons

22 of 39

SLIDE 23

Higgs mass measurement

Precise measurement of mH important for

determining couplings

For a shift in mass ∆mH = 400 MeV,

σ × BR(ZZ) changes by ≃ 3%

ATLAS mH measurement uses high

resolution modes H → γγ H → ZZ (∗) → 4ℓ

Improvements with respect to preliminary

results

Significantly improved e/γ calibration
Systematic on mH in γγ due to photon energy

scale reduced by factor 2.5

Improved lepton performace
Likelihood-based electron ID
E-p combination for electrons
S/B for 2µ2e final state improved from

1.2 → 1.8

Multivariate techniques in H → ZZ (∗) → 4ℓ
BDT as additional observable in fit → 8%

improvement compared to 1D

[GeV]

γ γ

m 110 120 130 140 150 160 weights - fitted bkg

∑

8
6
4
2

2 4 6 8

weights / GeV

∑

20 40 60 80 100 120 140 160 180 200 Data Combined fit: Signal+background Background Signal = 7 TeV s

1

Ldt = 4.5 fb

∫

= 8 TeV s

1

Ldt = 20.3 fb

∫

s/b weighted sum Mass measurement categories

ATLAS

[GeV]

l 4

m 80 90 100 110 120 130 140 150 160 170 Events / 2.5 GeV 5 10 15 20 25 30 35

Data = 1.66) µ = 124.5 GeV

H

Signal (m Background ZZ* t Background Z+jets, t Systematic uncertainty

l 4 → ZZ* → H

1

Ldt = 4.5 fb

∫

= 7 TeV: s

1

Ldt = 20.3 fb

∫

= 8 TeV: s

ATLAS 23 of 39

SLIDE 24

Higgs mass measurement

[GeV]

H

m 123 123.5 124 124.5 125 125.5 126 126.5 127 127.5 Λ

2ln

1 2 3 4 5 6 7

σ 1 σ 2

ATLAS

1

Ldt = 4.5 fb

∫

= 7 TeV s

1

Ldt = 20.3 fb

∫

= 8 TeV s l +4 γ γ Combined γ γ → H l 4 → ZZ* → H without systematics

Combined mass from a

simultaneous max. likelihood fit, where µγγ and µ4ℓ treated as independent free parameters

Individual measurements

compatibility ≃ 2.0σ

Compatibility in preliminary

result was 2.5σ

H → γγ : mH= 125.98 ± 0.42(stat) ± 0.28(sys) H → 4ℓ : mH= 124.51 ± 0.52(stat) ± 0.06(sys) Comined : mH= 125.36 ± 0.37(stat) ± 0.18(sys)

) [GeV]

l 4

m −

γ γ

= (m

H

m ∆

1
0.5

0.5 1 1.5 2 2.5 3 3.5 4 Λ

2ln

2 4 6 8 10 12

σ 1 σ 2 (0) Λ

2ln

σ 3

ATLAS

1

Ldt = 4.5 fb

∫

= 7 TeV s

1

Ldt = 20.3 fb

∫

= 8 TeV s

24 of 39

SLIDE 25

Measuring coupling properties

Most recent ATLAS couplings combination released March 2014
γγ, ZZ (∗) → 4ℓ, WW (∗)− → lνlν, τ +τ −, b¯

b

Also use combination to put constraints on new phenomena
Many of the results shown so far today not yet included in combination
Note measuring absolute couplings depends on total width:

σ × BR(i → H → f ) = σi · Γf ΓH

In SM ΓH ≃ 4 MeV!
Not possible to measure directly at the LHC
Alternatively, measure ratios of couplings
Dependence on ΓH cancels
Updated couplings combination with final results planned
Possibility to include searches for rare decays and t¯

tH production in future combinations

25 of 39

SLIDE 26

Production mode rates

ggF+ttH

µ /

VBF+VH

µ

1 2 3 4 5

ATLAS Prelim.

1

Ldt = 4.6-4.8 fb

∫

= 7 TeV s

1

Ldt = 20.3 fb

∫

= 8 TeV s

= 125.5 GeV

H

m

0.6

0.8

+

= 1.2

ggF+ttH

µ

VBF+VH

µ

γ γ → H

σ 1 σ 2 0.2

0.2

+ 0.2

0.4

+ 0.5

0.7

+

0.9

2.4

+

= 0.6

ggF+ttH

µ

VBF+VH

µ

4l → ZZ* → H

σ 1 0.2

0.3

+ 0.2

0.6

+ 0.9

2.3

+

1.0

1.9

+

= 1.8

ggF+ttH

µ

VBF+VH

µ

ν l ν l → WW* → H

σ 1 0.2

0.5

+ 0.4

1.3

+ 0.9

1.4

+

1.2

∞

+

= 1.7

ggF+ttH

µ

VBF+VH

µ

τ τ → H

0.3

∞

+ 0.6

∞

+ 1.0

5.3

+

0.5

0.7

+

= 1.4

ggF+ttH

µ

VBF+VH

µ

Combined

σ 1 σ 2 0.1

0.2

+ 0.2

0.4

+ 0.4

0.5

+

Total uncertainty σ 1 ± σ 2 ±

(stat.) σ

)

theory sys inc.

(

σ (theory) σ

No combination of µggF, µVBF

possible between decay modes

Can’t distinguish between production

and decay for deviations

Combine ratio instead

µVBF/µggF+ttH = 1.4+0.5

−0.4(stat)+0.4 −0.3(sys)

4.1σ evidence for VBF Higgs

production

ggF+ttH

µ /

VBF

µ

0.5

0.5 1 1.5 2 2.5 3 3.5 Λ

2 ln

2 4 6 8 10 12 14 16 18 20 22 24

combined SM expected

Preliminary ATLAS

1

Ldt = 4.6-4.8 fb

∫

= 7 TeV s

1

Ldt = 20.3 fb

∫

= 8 TeV s = 125.5 GeV

H

m

26 of 39

SLIDE 27

κ-framework

Framework for couplings based on LHC Higgs Cross Section Working Group

recommendations

Leading order framework for a single, SM-like Higgs boson under specific

assumptions:

Single resonance with a mass near 125 GeV
Zero width approximation holds
Tensor structure of couplings assumed to be the same as SM
JP = 0+
Define couplings scale factors κ :

σ · BR(i → H → f ) = σi · Γf ΓH = σSM

i

· ΓSM

f

ΓSM

H

· κ2

i · κ2 f

κ2

H

κi = 1 corresponds to the SM

⇒ Idea is to Look for deviations from SM rates

27 of 39

SLIDE 28

κ-framework

Framework makes no specific assumptions on additional states of new

physics which could interact with the state at ≃ 125 GeV, in particular on:

Additional Higgs bosons
Additional fermions, vector bosons or others scalars (which don’t acquire a VEV)
Invisible decay modes
Test benchmark scenarios based on this framework
Fermion vs vector couplings:
Tests EWSB, Yukawa coupling model
One scale factor for vector bosons and one for fermions
Fermion structure:
Many SM extensions (e.g. 2HDMs) predict deviations in fermion sector
One scale factor for up-type fermions and one for down-type
One scale factor for quarks and one for leptons
Several other benchmarks also tested

28 of 39

SLIDE 29

Vector boson vs fermion couplings: H → ZZ (∗) example

Benchmark model with one scale factor for all vector bosons (κV ), one for all

fermions (κF)

H Z Z κ2

V

κ2

f

g g

H Z Z κ2

V

κ2

V

W/Z W/Z q q µggF;H→ZZ = σ(ggF) · BR(H → ZZ) σSM (ggF) · BRSM (H → ZZ) → κ2

F · κ2 V

κ2

H(κ2 F , κ2 V )

µVBF;H→ZZ = κ2

V · κ2 V

κ2

H(κ2 F , κ2 V )

κH(κF, κV ) is a scale factor for Γtotal

H κ2

H(κ2 F , κ2 V ) = α · κ2 F + β · κ2 V

V

κ 0.6 0.8 1 1.2 1.4 1.6 1.8 2

F

κ 1 2 3 4 5

ATLAS l 4 → ZZ* → H

1

Ldt = 4.5 fb

∫

= 7 TeV s

1

Ldt = 20.3 fb

∫

= 8 TeV s = 125.36 GeV

H

m 68% CL 95% CL SM

29 of 39

SLIDE 30

Vector boson vs fermion couplings

Total width is sum of known SM Higgs decay modes
Modified appropriately with κV and κF

κV = 1.15 ± 0.08 κF = 0.99+0.17

−0.15

Only relative sign physical → set κV > 0
Sensitivity to relative sign from interference in

H → γγ decays

2D compatibility of SM with best fit 10%

Free parameters:

κV , κF

V

κ 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6

F

κ

2
1

1 2 3 4

bb → H bb → H τ τ → H τ τ → H 4l → H 4l → H ν l ν l → H ν l ν l → H γ γ → H γ γ → H

bb → H τ τ → H 4l → H ν l ν l → H γ γ → H Combined SM Best Fit

1

Ldt = 20.3 fb

∫

= 8 TeV s

1

Ldt = 4.6-4.8 fb

∫

= 7 TeV s

ATLAS Preliminary

V

κ 0.8 0.9 1 1.1 1.2 1.3 1.4 )

V

κ ( Λ

2 ln

2 4 6 8 10 ATLAS Preliminary

1

Ldt = 4.6-4.8 fb

∫

= 7 TeV, s

1

Ldt = 20.3 fb

∫

= 8 TeV, s b ,b τ τ ,ZZ*,WW*, γ γ → Combined H ]

F

κ ,

V

κ [

Observed SM expected F

κ

1
0.5

0.5 1 1.5 )

F

κ ( Λ

2 ln

2 4 6 8 10 ATLAS Preliminary

1

Ldt = 4.6-4.8 fb

∫

= 7 TeV, s

1

Ldt = 20.3 fb

∫

= 8 TeV, s b ,b τ τ ,ZZ*,WW*, γ γ → Combined H ]

F

κ ,

V

κ [

Observed SM expected

30 of 39

SLIDE 31

Vector boson vs fermion couplings

Assumption on total width gives strong constraint on κF
Total width in SM dominated by b, τ and gluon decay widths
Relax assumption by measuring ratios of scale factors
Take ratio of fermion and vector scale

factors λFV

Then κVV is an overall scale factor

which applies to all rates λFV = 0.86+0.14

−0.12

κVV = 1.28+0.16

−0.15

FV

λ

1.5
1
0.5

0.5 1 1.5 )

FV

λ ( Λ

2 ln

2 4 6 8 10

ATLAS Preliminary

1

Ldt = 4.6-4.8 fb

∫

= 7 TeV, s

1

Ldt = 20.3 fb

∫

= 8 TeV, s b ,b τ τ ,ZZ*,WW*, γ γ → Combined H ]

VV

κ ,

FV

λ [

Observed SM expected

Free parameters:

λFV = κF/κV , κVV = κV · κV /κH

31 of 39

SLIDE 32

Up-type vs down-type fermions

One scale factor for up-type fermions and one for down-type
Some SM extensions (e.g some 2HDMs) predict different couplings for up- and

down-type fermions

e.g. MSSM
Take ratio of down and up scale factors λdu

λdu = 0.95+0.20

−0.18∗

λVu = 1.21+0.24

−0.26

κuu = 0.86+0.41

−0.21 For positive minima

Little sensitivity to relative sign
3D compatibility with SM 20%
3.6σ evidence for coupling to

down-type fermions

du

λ

2
1.5
1
0.5

0.5 1 1.5 2 )

du

λ ( Λ

2 ln

2 4 6 8 10

ATLAS Preliminary

1

Ldt = 4.6-4.8 fb

∫

= 7 TeV, s

1

Ldt = 20.3 fb

∫

= 8 TeV, s b ,b τ τ ,ZZ*,WW*, γ γ → Combined H ]

uu

κ ,

du

λ ,

Vu

λ [

Observed SM expected

Free parameters:

λdu = κd/κu, λVu = κV /κu, κuu = κu · κu/κH

32 of 39

SLIDE 33

Off shell Higgs couplings

H → VV high mass region has sensitivity to
ff-shell Higgs production

dσpp→H→ZZ dM2

4ℓ

∼ g2

Hgg g2 HZZ

(M2

4ℓ − m2 H) + m2 HΓ2 H

Using κ language

µon−shell = κ2

g κ2 Z

κ2

H

µoff −shell = κ2

g κ2 Z

[GeV]

4l

m 200 400 600 800 1000 [fb/GeV]

4l

/dm σ d

6

10

5

10

4

10

3

10

2

10

1

10

ZZ (S) → H* → gg ZZ (B) → gg ) ZZ → (H* → gg =10)

ff-shell

µ ) ZZ ( → (H* → gg

Simulation Preliminary

ATLAS

µ 2 e 2 → ZZ → gg = 8 TeV s H SM

Γ /

H

Γ 2 4 6 8 10 12 14 Λ

2ln

2 4 6 8 10 12 14

Preliminary

ATLAS

combined

n-shell

l +4 l +4 ν 2 l 2

1

Ldt = 20.3 fb

∫

= 8 TeV: s expected with syst. expected no syst.

bserved
Combining on- and off-shell results, can

interpret as measurement of ΓH

Measurement performed by ATLAS using

H → ZZ (∗) → 4ℓ and H → ZZ (∗) → 2ℓ2ν

ΓH

ΓSM

H

< 5.7 at 95% CL

33 of 39

SLIDE 34

Constraints on new phenomena I: Additional Electroweak singlet

Two Higgs bosons, one light (h), one heavy(H)
Couple to vector bosons and fermions similar to SM but modified by scale

factors

κ + κ′ = 1
h couplings same as SM, modified by κ
H couplings modified to take into account new decay modes (e.g. H → hh)

µH = κ′2(1 − BRH,new) κ′2 = 1 − µh

Best fit at κ′2 = −0.30+0.17

−0.18

1.5σ from physical boundary

κ′2 ≥ 0

Set limits in µH, BRH,new

plane

H

µ

=0.1

2

’ κ =0.2

2

’ κ = . 3

2

’ κ = . 4

2

’ κ = . 5

2

’ κ =0.6

2

’ κ = . 7

2

’ κ =0.8

2

’ κ = . 9

2

’ κ =1.0

2

’ κ =0.1

H,SM

Γ /

H

Γ = . 5

H , S M

Γ /

H

Γ =1.0

H,SM

Γ /

H

Γ =5.0

H , S M

Γ /

H

Γ =100

H,SM

Γ /

H

Γ

ATLAS Preliminary

1

Ldt = 4.6-4.8 fb

∫

= 7 TeV: s

1

Ldt = 20.3 fb

∫

= 8 TeV: s b ,b τ τ ,ZZ*,WW*, γ γ → Combined h EW singlet

Obs. 95% CL
Exp. 95% CL

SM

0.2 0.4 0.6 0.8 1

H,new

BR 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

34 of 39

SLIDE 35

Constraints on new phenomena II: Invisible branching ratio and dark matter portals

Derive upper limits on Higgs BR to invisible

final states

Uses couplings combination combined with

upper limits on ZH → ℓℓ + E miss

T

process

BRi < 0.37 at 95% CL

i

BR

1
0.8 -0.6 -0.4 -0.2

0.2 0.4 0.6 0.8 1 )

i

(BR Λ

2 ln

2 4 6 8 10 12 14 Preliminary ATLAS

1

Ldt = 4.6-4.8 fb

∫

= 7 TeV, s

1

Ldt = 20.3 fb

∫

= 8 TeV, s ]

i

, BR

g

κ ,

γ

κ [ : b , b τ τ , ZZ*, WW*, γ γ → h , b , b τ τ , ZZ*, WW*, γ γ → h :

T miss

ll + E → Zh

bs.

exp.

bs.

exp.

[GeV]

χ

m 1 10

2

10

3

10 ]

2

[cm

N

χ

σ

57

10

55

10

53

10

51

10

49

10

47

10

45

10

43

10

41

10

39

10

DAMA/LIBRA (99.7% CL) CRESST (95% CL) CDMS (95% CL) CoGeNT (90% CL) XENON10 (90% CL) XENON100 (90% CL) LUX (95% CL) Scalar WIMP Majorana WIMP Vector WIMP

ATLAS Preliminary

Higgs portal model: ATLAS (95% CL) in

1

dt = 4.6-4.8 fb L

∫

= 7 TeV, s

1

dt = 20.3 fb L

∫

= 8 TeV, s , ν l ν l → WW* → 4l, h → ZZ* → , h γ γ → h

miss T

ll+E → bb, Zh → , h τ τ → h

Higgs portal models introduce

weakly-interacting massive particles as dark matter candidates

Assumed to interact weakly with SM particles

except Higgs boson

Can compare limits with direct dark matter

searches

Assuming mWIMP < 0.5 · mH and

H → 2WIMPs accounts for all of BRi

35 of 39

SLIDE 36

LHC upgrade timescale

HL-LHC upgrade proposed
Goal to collect 3000 fb−1 by 2035
Corresponding proposals for upgrades of the LHC experiments
Central feature of ATLAS upgrade programme a new, all silicon tracking system

36 of 39

SLIDE 37

Prospects for Higgs coupling measurements at a HL-LHC

µ / µ ∆ 0.2 0.4

(comb.) (VBF-like) (comb.) (incl.) (comb.) (comb.) (comb.)

ATLAS Simulation Preliminary

= 14 TeV: s

1

Ldt=300 fb

∫

;

1

Ldt=3000 fb

∫

γ γ → H ZZ → H WW → H γ Z → H b b → H τ τ → H µ µ → H γ γ → H ZZ → H WW → H γ Z → H b b → H τ τ → H µ µ → H

ATLAS has studied the prospects for Higgs

coupling studies with 3000 fb−1

Generator-level MC with parameterised model

for detector efficiency and resolution

Parameterisations from Geant4 simulation
140 interactions per bunch crossing
Systematic uncertainties same as run I
Data-driven uncertainties scaled with int lumi
Hashed bands: theoretical uncertainties at

their current level

Projections typically based on older versions of

analyses - do not include recent improvements

Possible to measure decay rates to sub 10%

level

37 of 39

SLIDE 38

Prospects for Higgs coupling measurements at a HL-LHC

)

Y

κ

X

κ ( ∆ =

XY

λ ∆ 0.05 0.1 0.15 0.2 0.25

)Z γ (Z

λ

Z γ

λ

gZ

λ

Z µ

λ

Z τ

λ

bZ

λ

tg

λ

WZ

λ

gZ

κ

ATLAS Simulation Preliminary

= 14 TeV: s

1

Ldt=300 fb

∫

;

1

Ldt=3000 fb

∫

Potential to measure coupling ratios down to

few % level with 3000 fb−1

Projections in terms of scaling of couplings as

for run I, but likely to move to a more general framework, e.g. effective field theory

i

y

3

10

2

10

1

10 1 Z W t b τ µ ATLAS Simulation Preliminary

= 14 TeV s

ν l ν l → WW* → 4l, h → ZZ* → , h γ γ → h γ Z → , h µ µ → bb, h → , h τ τ → h ]

µ

κ ,

τ

κ ,

b

κ ,

t

κ ,

W

κ ,

Z

κ [ =0

i,u

BR

1

dt = 300 fb L

∫

1

dt = 3000 fb L

∫

[GeV]

i

m

1

10 1 10

2

10 Ratio to SM

0.8 0.9 1 1.1 1.2

38 of 39

SLIDE 39

Conclusion

Parameter value

2
1

1 2

ATLAS Preliminary

1

Ldt = 4.6-4.8 fb

∫

= 7 TeV s

1

Ldt = 20.3 fb

∫

= 8 TeV s

= 125.5 GeV

H

m

0.08

0.08

+

=1.15

V

κ

σ 1 σ 2

F

κ ,

V

κ Model: =10%

SM

p 0.15

0.17

+

=0.99

F

κ

σ 1 σ 2 0.12

0.14

+

=0.86

FV

λ

σ 1 σ 2

VV

κ ,

FV

λ Model: =10%

SM

p 0.29

0.14

+

=0.94

WZ

λ

σ 1 σ 2

ZZ

κ ,

FZ

λ ,

WZ

λ Model: =19%

SM

p

[0.78,1.15] ∪ [-1.24,-0.81] ∈

du

λ

σ 1 σ 2

uu

κ ,

Vu

λ ,

du

λ Model: =20%

SM

p

[0.99,1.50] ∪ [-1.48,-0.99] ∈

lq

λ

σ 1 σ 2

qq

κ ,

Vq

λ ,

lq

λ Model: =15%

SM

p 0.13

0.15

+

=1.08

g

κ

σ 1 σ 2

γ

κ ,

g

κ Model: =9%

SM

p 0.12

0.15

+

=1.19

γ

κ

σ 1 σ 2

i,u

, B

γ

κ ,

g

κ Model: =18%

SM

p 0.30

0.29

+

=-0.16

i.,u.

BR

σ 1 σ 2

<0.41

i.,u.

BR @ 95% CL

Total uncertainty σ 1 ± σ 2 ±

So far no significant

deviation from SM

Increased precision

anticipated during next LHC runs and beyond

ATLAS used LHC run I dataset to probe the

coupling properties of the Higgs

Results suggest that a non-zero VEV of a

scalar doublet is indeed responsible for EWSB

Evidence for Higgs decays to fermions also

seen in ττ final state

Observed rates agree with SM Yukawa coupling

prediction

[GeV]

H

M

60 70 80 90 100 110 120 130 140

2

χ ∆

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 σ 1 σ 2

measurement

H

SM fit with M measurement

H

SM fit w/o M ATLAS measurement [arXiv:1406.3827] CMS measurement [arXiv:1407.0558]

G fitter SM

Jul ’14

2

χ ∆ 4.5 5

measurement

H

SM fit with M

G fitter SM

Jul ’14

39 of 39