[PPT] - Tevatron measurements on Standad Model Higgs 49 th Rencontres de PowerPoint Presentation

SLIDE 1

Tevatron measurements on Standad Model Higgs

49th Rencontres de Moriond Electroweak Interactions and Unified Theories

Federico Sforza

n behalf of the CDF and D0 Collaborations

Max Planck Institut für Physik

20th March 2014 - La Thuile

F. Sforza (Max Planck Institut für Physik)

20th March 2014 1 / 19

SLIDE 2

The Tevatron

Presented analyses use full Run II dataset: Up to

L ≃ 10 fb−1 of p¯

p collisions at √s = 1.96 TeV per experiment (≃ 12 fb−1 delivered) 1km p p

store number

1000 2000 3000 4000 5000 6000 7000 8000 9000 50 100 150 200 250 300 350 400 450

01/11 01/10 01/09 01/08 01/07 01/06 01/05 01/04 01/03

)

1

s

2

cm

30

10 × Initial Luminosity (

Store initial luminosity ×20 increase over years ⇒ driven by abundance of anti-protons

Tevatron facts: Two instrumented collision points: CDF & D0 experiments First superconducting accelerator and largest anti-matter source in the world Run I and Run II cover almost 20 years of physics

F. Sforza (Max Planck Institut für Physik)

20th March 2014 2 / 19

SLIDE 3

The Tevatron

A Very Brief History of Higgs Searches

)

2

(GeV/c

H

m 100 110 120 130 140 150 160 170 180 190 200 95% CL Limit / SM 5 10 15 20 25 30 35 40 DØ Expected CDF Expected Tevatron Expected Tevatron Observed

Tevatron Run II Preliminary

1

Ldt=0.3-1.0 fb

∫

LEP Excluded

Tevatron role was unexpected several years ago: First Tevatron combination for SM Higgs 95% C.L.: ⇒ 2006, 0.3 − 1 fb −1 (CDF 8384 & D0 5227 Notes) 95% exclusion sensitivity O(10)×SM 100 fb−1 needed to reach 2σ sensitivity!

F. Sforza (Max Planck Institut für Physik)

20th March 2014 3 / 19

SLIDE 4

The Tevatron

A Very Brief History of Higgs Searches

)

2

(GeV/c

H

m 100 110 120 130 140 150 160 170 180 190 200 95% CL Limit / SM 5 10 15 20 25 30 35 40 DØ Expected CDF Expected Tevatron Expected Tevatron Observed

Tevatron Run II Preliminary

1

Ldt=0.3-1.0 fb

∫

LEP Excluded

Tevatron role was unexpected several years ago: First Tevatron combination for SM Higgs 95% C.L.: ⇒ 2006, 0.3 − 1 fb −1 (CDF 8384 & D0 5227 Notes) 95% exclusion sensitivity O(10)×SM 100 fb−1 needed to reach 2σ sensitivity! The July 2012 discovery of a new particle compatible with SM Higgs (mH ≈ 125 GeV/c2):

ATLAS, CMS observation in 4ℓ, γγ final states:

[GeV]

H

m 110 115 120 125 130 135 140 145 150 Local p

11

10

10

10

9

10

8

10

7

10

6

10

5

10

4

10

3

10

2

10

1

10 1 Obs. Exp. σ 1 ±

1

Ldt = 5.8-5.9 fb

∫

= 8 TeV: s

1

Ldt = 4.6-4.8 fb

∫

= 7 TeV: s

ATLAS 2011 - 2012

σ σ 1 σ 2 σ 3 σ 4 σ 5 σ 6

Phys. Lett. B 716 (2012) 1 and 30

Tevatron evidence in b¯ b final state:

)

2

(GeV/c

H

m 100 105 110 115 120 125 130 135 140 145 150 Background p-value

3

10

2

10

1

10 1 10

2

10

σ 1 σ 2 σ 3

1

9.7 fb ≤

int

Tevatron Run II, L

Observed

b

1-CL Expected

b

1-CL 1 s.d. ± 2 s.d. ±

1

9.7 fb ≤

int

Tevatron Run II, L

Phys.Rev.Lett. 109, 071804 (2012)

F. Sforza (Max Planck Institut für Physik)

20th March 2014 3 / 19

SLIDE 5

The Tevatron

Higgs Production Mode Differences at Tevatron and LHC

Tevatron Higgs production modes (1.96 TeV): LHC Higgs production modes (8 TeV):

[GeV]

H

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

2

10

1

10 1 10

2

10 = 8 TeV s

LHC HIGGS XS WG 2012 H (NNLO+NNLL QCD + NLO EW) → pp q q H ( N N L O Q C D + N L O E W ) → p p WH (NNLO QCD + NLO EW) → pp ZH (NNLO QCD +NLO EW) → pp ttH (NLO QCD) → pp

Higgs production rate at LHC much higher than at Tevatron: LHC gluon fusion × 20, Vector Boson Fusion (VBF) × 30 ⇒ Abundant production modes for analysis of clean final states with small BR (γγ, ZZ, WW, ττ) LHC VH associate production × 4, also higher background: ⇒ Relevant and complementary studies from Tevatron!

F. Sforza (Max Planck Institut für Physik)

20th March 2014 4 / 19

SLIDE 6

SM Higgs Analyses: Cross Section and Couplings

Analysis Classification

SM Higgs decay branching fractions:

High Mass: High BR final states for mH 135 GeV/c2 Main channel: gg → H → WW WW → ℓνℓν: low background High Higgs production rate Low Mass: High BR final states for mH 135 GeV/c2 main channel: qq → VH → b¯ b V leptonic decay: online selection and background reduction Peculiarity of mH = 125 GeV/c2: BR(H → b¯ b) ≃ 0.58, BR(H → WW) ≃ 0.21 ⇒ Both low and high mass analyses contribute to properties study

F. Sforza (Max Planck Institut für Physik)

20th March 2014 5 / 19

SLIDE 7

SM Higgs Analyses: Cross Section and Couplings

Overview of H → WW Analysis

Lepton plus ✁ ET selection (also hadronic τ): ⇒ s/b event categorization to enhance sensitivity Low Higgs mass resolution because of 2ν Lepton kinematic correlation for MVA discriminants: ⇒ Boosted Decision Trees (BDT), usually for D0 ⇒ Neural Networks (NN), usually for CDF

Different di-lepton kinematic of H → WW decay and WW EWK production (background): Example of CDF H → WW NN output: Phys.Rev. D 88, 052012 (2013) Example of D0 H → WW BDT output: (s/b)

10

log

3.5
3
2.5
2
1.5
1
0.5

Events/0.30

200 400 600 800 1000 1200 1400 1600 1800 Data 20 × Signal Z+jets Diboson W+jets Multijet t t

Bkg. syst.

T

E , ll +

1

DØ, 9.7 fb Phys.Rev. D. 88, 052006 (2013)

2008: first post LEP analysis to exclude presence SM Higgs boson in mass range

F. Sforza (Max Planck Institut für Physik)

20th March 2014 6 / 19

SLIDE 8

SM Higgs Analyses: Cross Section and Couplings

Overview of Low Mass Higgs Analyses

VH → b¯ b is the most sensitive channel at Tevatron: 3 analyses with similar topology: leptons (charged or neutral) + heavy flavor jets

p¯ p → WH → ℓν + b¯ b p¯ p → ZH → ℓℓ + b¯ b p¯ p → VH → νν(ν✄ ℓ) + b¯ b

CDF: Phys.Rev.Lett. 109, 111804 (2012), D0: Phys.Rev.Lett. 109, 121804 (2012), Phys.Rev.D. 88, 052008 (2013) CDF: Phys.Rev.Lett. 109, 111803 (2012), D0: Phys. Rev. Lett. 109, 121803 (2012), Phys.Rev.D. 88, 052010 (2013) CDF: Phys.Rev.D 87, 052008 (2013), D0: Phys.Lett.B 716, 285 (2012)

Similar background sources, different relative fractions between final states Aim to identify Higgs Mb¯

b resonance over falling background

Background estimate with mixture of MC/data driven extraction from Control Regions (CR) Backgrounds Shape Normalization WW,WZ,ZZ, t¯ t, single-top MC based NLO,NNLO (Theory) Multi-Jet (MJ) Data driven Fit to data W/Z+light flavor MC (ALPGEN), CR weighted data CR W/Z+heavy flavor MC based (ALPGEN) LO, fit to data Statistically limited dataset ⇒ Relax cuts ⇒ Keep background under control ⇒ Iterate

F. Sforza (Max Planck Institut für Physik)

20th March 2014 7 / 19

SLIDE 9

SM Higgs Analyses: Cross Section and Couplings

Low Mass Searches Optimization

Every aspect of analyses thoroughly optimized (often using MVA): Inclusive trigger strategy: single lepton, only✚ ET , multiple objects (

✚

ET +jets) Improved ℓ/✁ ET offline ID: relaxed cuts increases MJ ⇒ improve lepton ID/MJ-rejection b-tag: reduce background to 1/100 but limits jet selection efficiency (≃ 50%) Final Discriminant: large irreducible backgrounds ⇒ MVA sensitivity increase by 10-20% MVA b-tagging for both D0 and CDF Tunable efficiency/contamination working point Maximize significance from s/b categorization of signal region

D0 NN b-tagger output composition:

V. M. Abazov et al., Nucl.Instrum.Methods Phys.Res., Sect. A 620, 490 (2010)

CDF b-tag working point comparison:

J. Freeman et al., Nucl.Instrum.Methods Phys.Res., Sect. A 697, 64 (2013)

D0 b-tag algorithm improvements and working points:

F. Sforza (Max Planck Institut für Physik)

20th March 2014 8 / 19

SLIDE 10

SM Higgs Analyses: Cross Section and Couplings

Low Mass Searches Optimization

Every aspect of analyses thoroughly optimized (often using MVA): Inclusive trigger strategy: single lepton, only✚ ET , multiple objects (

✚

ET +jets) Improved ℓ/✁ ET offline ID: relaxed cuts increases MJ ⇒ improve lepton ID/MJ-rejection b-tag: reduce background to 1/100 but limits jet selection efficiency (≃ 50%) Final Discriminant: large irreducible backgrounds ⇒ MVA sensitivity increase by 10-20% Single channel sensitivity improvements w.r.t. 1 fb−1 analysis was also > 200% (over luminosity)!

F. Sforza (Max Planck Institut für Physik)

20th March 2014 8 / 19

SLIDE 11

SM Higgs Analyses: Cross Section and Couplings

Analysis Validation with VZ → HF

Important analysis validation using known SM process as signal VZ associate production in s−channel with Z → b¯ b mimics VH → b¯ b signature σVZ × BR(Z → HF) about 6 times VH (MH = 125) Higher background due to W+jets Mjj spectrum ⇒ very small s/b and challenging measurement!

WZ → HF evidence

CDF and D0 low mass analyses combined looking at VZ → HF signal Same data-set, analysis techniques, MVA discriminant strategy ⇒ σVZ = 3.0 ± 0.6(stat) ± 0.7(syst) pb ⇒ Strong signal evidence at 4.6 σ ⇒ Consistent with σSM,NLO

VZ

= 4.4 ± 0.3 pb

)

2

Dijet Mass (GeV/c 50 100 150 200 250 300 350 400 )

2

Events / (20 GeV/c 200 400 600 800

Data — Bkgd WZ ZZ Higgs Signal

2

=125 GeV/c

H

m

1

10 fb ≤

int

Tevatron Run II, L 1+2 b-Tagged Jets Background subtracted di-jet invariant mass

F. Sforza (Max Planck Institut für Physik)

20th March 2014 9 / 19

SLIDE 12

SM Higgs Analyses: Cross Section and Couplings

H → b¯ b Results

Phys.Rev.Lett. 109, 071804 (2012): H → b¯ b low mass VH Tevatron combination in 2012 Phys.Rev.D 88, 052014 (2013): properties of H → b¯ b from all channel update in July 2013

SM Higgs 95% exclusion limits Cross section × BR measurement

Significant excess over background only hypothesis: Analysis using both Log Likelihood Ratio (LLR) and Bayesian posterior cross section measurement σ(WH + ZH) × BR(H → b¯ b) = 0.19 ± 0.09 pb (SM exp. 0.12 ± 0.01) ⇒ Measurement of H → b¯ b competitive with LHC

F. Sforza (Max Planck Institut für Physik)

20th March 2014 10 / 19

SLIDE 13

SM Higgs Analyses: Cross Section and Couplings

Tevatron Combination for All Analysis Channels

Next step ⇒ analyze SM Higgs from combination of all analysis channels, > 100! H → WW and H → b¯ b are the most important, but also H → γγ, H → ττ SM Higgs hypothesis testing is possible only looking at all the predicted decay modes Measure parameter of new particle: Summer 2013 Results The global picture: Classification of all final discriminants in s/b bins Preserve importance of each data event log10(s/b) shows agreement

ver 5 orders of magnitude

10-4 10-3 10-2 10-1 1 10 102 103 104 105 106

4
3
2
1

log10(s/b) Events/0.16

Data Background fit SM Higgs signal mH=125 GeV/c2 Tevatron Run II, Lint ≤ 10 fb-1 SM Higgs combination

F. Sforza (Max Planck Institut für Physik)

20th March 2014 11 / 19

SLIDE 14

SM Higgs Analyses: Cross Section and Couplings

p − values and Cross Section

The full Tevatron combination results:

Analysis of each channel discriminant with combined likelihood p-value 3.0σ (local) at mH = 125 GeV/c2 (1.9 expected) µ = σobs/σSM = 1.44+0.59

−0.56 at mH = 125 GeV/c2

Consistent cross section between channels and with SM expectation

)

2

(GeV/c

H

m 100 120 140 160 180 200 Background p-value

5

10

4

10

3

10

2

10

1

10 1 10

2

10

3

10

σ 1 σ 2 σ 3 σ 4

1

10 fb ≤

int

Tevatron Run II, L SM Higgs Combination

Observed Expected w/ Higgs 1 s.d. ± Expected 2 s.d. ± Expected )

2

=125 GeV/c

H

1.0 (m ×

H

σ )

2

=125 GeV/c

H

1.5 (m ×

H

σ

0.5 1 1.5 2 2.5 3 3.5 4 100 120 140 160 180 200 mH (GeV/c2) σ/SM SM=1 Tevatron Run II, Lint ≤ 10 fb-1 SM Higgs combination ± 1 s.d. ± 2 s.d. Observed σH x 1.5 (mH=125 GeV/c2) σH x 1.0 (mH=125 GeV/c2)

F. Sforza (Max Planck Institut für Physik)

20th March 2014 12 / 19

SLIDE 15

SM Higgs Analyses: Cross Section and Couplings

Higgs Coupling Measurements

Most sensitive Higgs production and decay modes via W, Z, b-quark Extract coupling deviations from SM prediction from per-channel signal rates Assumptions: mH = 125 GeV/c2, negligible width, CP 0+, no invisible decay

κf fermion couplings scale κV boson coupling scale (if κZ ≡ κW ) Examples: → Γb¯

b ∝ κ2 f , Γτ ¯ τ ∝ κ2 f , ΓZZ ∝ κ2 V

→ ΓWW ∝ Rκ2

V (with R = κW /κZ )

Preserve unitary in BR:

es. Γγγ ∝ (1.28κf − 0.28κV )2

Study of coupling multiplicative parameters with 1-dim and 2-dim (shown) Bayesian posteriors κf vs κV (if κW = κZ )

F. Sforza (Max Planck Institut für Physik)

20th March 2014 13 / 19

SLIDE 16

SM Higgs Analyses: Cross Section and Couplings

Vector Boson Couplings: κW vs κZ

Separate measurement of κZ vs κW κf is marginalized 95% C.L. exclusion of no-Higgs hypothesis: (κZ, κW) = (0, 0) 2-dim best fit: ⇒ (κW, κZ) = (1.25, ±0.90)

WZ

θ

0.5 1 1.5 Posterior probability density 0.5 1 1.5

floating

f

κ

Best Fit SM=π/4 68% C.L. 95% C.L.

10 fb ≤

int

Tevatron Run II, L

Test of SU(2) custodial symmetry:

by measuring λWZ = κW/κZ θWZ = tan−1(κZ/κW) = tan−1(1/λWZ) θWZ = 0.68+0.21

−0.41 ⇒ λWZ = 1.24+2.34 −0.42

F. Sforza (Max Planck Institut für Physik)

20th March 2014 14 / 19

SLIDE 17

Additional Studies: Spin, Parity, and Exotic Final States Analyses

Additional Properties: Spin and Parity

VH production processes differ depending on JP assignment Kinematic differences from behaviors at production threshold, if β = 2p/√s: 0+: S-wave production, σ ∝ β near threshold 0−: P-wave production, σ ∝ β3 near threshold 2+: D-wave dominates for graviton-like coupling, σ ∝ β5

cf. Ellis, et al., JHEP 1211, 134 (2012),

and Miller, et al., PLB 505, 149, (2001)

Probe Higgs JP with VH total mass variables ⇒ background discrimination better than for 0+: 2+ result Summer 2013: D0 Note 6387 0− recent result (Autumn 2013): D0 Note 6406

F. Sforza (Max Planck Institut für Physik)

20th March 2014 15 / 19

SLIDE 18

Additional Studies: Spin, Parity, and Exotic Final States Analyses

Results of Spin and Parity Analysis

Known mH used in analysis optimization: ⇒ selection of High/Low purity regions in reconstructed mass LLR test statistics used to distinguish two hypothesis with CLs: ⇒ H1 (Test): background plus 0− or 2+ Higgs signal ⇒ H0 (Null): background plus 0+ Higgs signal Two scenarios: SM like σVH × BR(b¯ b) (shown), or D0 measured rate (µ = 1.23)

LLR

60
40
20

20 40 60 Pseudoexperiments 500 1000 1500 2000 2500 3000 LLR + σ 1 ± LLR + σ 2 ± LLR + LLR

Observed LLR

DØ Preliminary

1

<= 9.7 fb =1.00 µ LLR

60
40
20

20 40 60 Pseudoexperiments 1000 2000 3000 4000 5000 6000 LLR

+

σ 1 ± LLR

+

σ 2 ± LLR

+

LLR

+

2 Observed LLR

1

DØ Preliminary, 9.7 fb =1.00 µ

Exclusions results:

0− model excluded at > 97.9% C.L. (2.3σ obs, 3.1σ exp) 2+ model excluded at > 99.9% C.L. (2.4σ obs, 3.2σ exp) Very good sensitivity but model dependent assumptions

F. Sforza (Max Planck Institut für Physik)

20th March 2014 16 / 19

SLIDE 19

Additional Studies: Spin, Parity, and Exotic Final States Analyses

Invisible Higgs

New CDF analysis testing exotic Higgs models: CDF Note 11068 First measurement at the Tevatron probing σZH × BR(H → invisible) If kinematically accessible, decays to heavy unknown and weakly interacting particles are likely because of Yukawa coupling (∝ m) to Higgs ZH → ℓℓ+✁ ET : 82 < Mℓℓ < 100 GeV/c2, clean signature Bayesian 95% C.L. exclude 100% BR(H → invisible) for mH < 120 GeV/c2

F. Sforza (Max Planck Institut für Physik)

20th March 2014 17 / 19

SLIDE 20

Conclusions and Prospects

Summary and Prospects

Summary of Tevatron Run II Higgs studies:

High and low Higgs mass analysis channels based

n full dataset completed and published

Data show a consistent picture of the SM Higgs: ⇒ p-value 3.0σ (local) at mH = 125 GeV/c2 ⇒ µ = σobs/σSM = 1.44+0.59

0.56

at mH = 125 GeV/c2

Recent results and work in progress:

VH → b¯ b datasets re-analyzed by D0 collaboration extracting spin and CP measurements: ⇒ JP = 0+ nature of the new particle are favored ⇒ JP = 2+, 0− models rejected at > 97% C.L. ⇒ Tevatron combination is in progress Exotic Higgs properties: ⇒ new analysis H →invisible

(σ × BR)/SM for decay channel:

Br)/SM × σ Best Fit (

1 2 3 4 5 6 7 8 9 10

b Vb → VH

τ

+

τ → H

W

+

W → H γ γ → H

Combined (68% C.L.) Single channel

1

10 fb ≤

int

Tevatron Run II, L mH=125 GeV/c2

Tevatron analyses still provide good sensitivity to H → b¯ b final state and Higgs properties studies are often complementary to LHC

F. Sforza (Max Planck Institut für Physik)

20th March 2014 18 / 19

SLIDE 21

Conclusions and Prospects

Thanks for Your Attention

F. Sforza (Max Planck Institut für Physik)

20th March 2014 19 / 19

SLIDE 22

BACK UP

Back Up Slides

F. Sforza (Max Planck Institut für Physik)

20th March 2014 20 / 19

SLIDE 23

BACK UP

All Channels and Analysis Details References

www-cdf.fnal.gov/physics/new/hdg/Results.html www-d0.fnal.gov/Run2Physics/WWW/results/higgs.htm tevnphwg.fnal.gov/results/SM_Higgs_Summer_13

F. Sforza (Max Planck Institut für Physik)

20th March 2014 21 / 19

SLIDE 24

BACK UP

The CDF and D0 Experiments

Multipurpose detectors:

Silicon (|η| < 2.5, r ≃ 20 cm) Inner Tracker Silicon (|η| < 3.0, r ≃ 10 cm) Drift cell (|η| < 1.1, r ≃ 130 cm) Outer Tracker Fiber (|η| < 1.7, r ≃ 50 cm) Pb/Fe/Scintillators (|η| < 3.6) Calorimeters LAr/U (|η| < 4.0) Drift/Scintillators (|η| < 1.5) Muon Chambers Drift/Scintillators |η| < 2.0

F. Sforza (Max Planck Institut für Physik)

20th March 2014 22 / 19

SLIDE 25

BACK UP

Analysis Improvement Examples

Low trigger thresholds (multi-dim. turn-on model): Lepton ID with NN selection: Variety of MJ-rejection techniques (here cut on SVM): Z → ℓℓ kinematic fit to improve Mjj resolution:

F. Sforza (Max Planck Institut für Physik)

20th March 2014 23 / 19

SLIDE 26

BACK UP

CDF ZH → νν + HF Update

2012 result documented in Phys. Rev. Lett. 109, 111805 (2012), Updated 2013 result documented in Phys.Rev.D 87, 052008 (2013) Different b-tagging and, therefore, different signal region categorization: new result more sensitive but with lower observed limit Fluctuation possible with 7% probability tested with P .E.

Two-sided p-value by calculating the conditional probability of obtaining a HOBIT result that is as or more discrepant than what we observe, given the S-J reanalysis observed limit

F. Sforza (Max Planck Institut für Physik)

20th March 2014 24 / 19

SLIDE 27

BACK UP

SM Higgs Compatibility Between Final States

(σH × BR)/SM in different final states:

F. Sforza (Max Planck Institut für Physik)

20th March 2014 25 / 19

SLIDE 28

BACK UP

1-Dim κf, κW, and κZ

f

κ

5

5 Posterior density 0.1 0.2 0.3 0.4

SM=1 Local maxima 68% C.L. 95% C.L.

1

10 fb ≤

int

Tevatron Run II, L

=1

Z

κ =

W

κ W

κ

2
1

1 2 Posterior density 0.2 0.4 0.6 0.8 1

SM=1 Local maxima 68% C.L. 95% C.L.

1

10 fb ≤

int

Tevatron Run II, L

=1

f

κ =

Z

κ Z

κ

2
1

1 2 Posterior density 0.1 0.2 0.3 0.4 0.5 0.6

SM=1 Local maxima 68% C.L. 95% C.L.

1

10 fb ≤

int

Tevatron Run II, L

=1

f

κ =

W

κ

F. Sforza (Max Planck Institut für Physik)

20th March 2014 26 / 19

SLIDE 29

BACK UP

Spin Exclusions Using Measured Higgs Signal Strenght

F. Sforza (Max Planck Institut für Physik)

20th March 2014 27 / 19