[PPT] - Constraining the top Yukawa coupling from tt differential PowerPoint Presentation

SLIDE 1

Young Scientist Forum 54th Rencontres de Moriond La Thuile, Italy, 16-23 March 2019

Constraining the top Yukawa coupling from tt differential distributions in lepton+jets at 13 TeV

Yi-Ting Duh

(Brown University & University of Rochester)

n behalf of the CMS Collaboration

CMS-PAS-TOP-17-004

SLIDE 2

Corrections due to electroweak boson exchange, including the Higgs boson,

between the final state top quarks

Have a very small effect on the total cross section; Not implemented in the MC

event generators

However, they have a sizable impact on differential distributions and sensitive

to the top quark Yukawa coupling through weak force mediated corrections, especially near the energy of production threshold [Ref].

Yukawa coupling in tt differential

Yi-Ting Duh

2

Example diagrams of tt production and the virtual Higgs exchange

In this analysis: explore a complementary approach to measure top quark Yukawa

coupling by utilizing precise measurements of tt differential distributions Ref: Phys. Rev. D 91 (2015) 014020

SLIDE 3

Yukawa coupling in tt differential

Yi-Ting Duh

3

Yt : the ratio of the top quark Yukawa coupling to its SM predicted value
Calculate weak correction factors for different value of Yt in a given (Mtt, Δytt)
Apply the correction factors at the parton level to existing tt simulated samples.

Obtain distributions at detector level that can be directly compared to data.

Invariant mass of top pairs Rapidity difference between top and antitop

[GeV]

t t

M

400 600 800 1000 1200 1400 1600 1800 2000

)

t t

/dM

LO

σ )/(d

t t

/dM

EW

σ δ (d

0.1

0.1 0.2 0.3 0.4 0.5

= 0

t

Y = 1

t

Y = 2

t

Y = 4

t

Y top mass uncert

CMS

Preliminary Simulation

The most sensitive region

(HATHOR inputs)

t t

y Δ

3
2
1

1 2 3

)

t t

y Δ /d

LO

σ )/(d

t t

y Δ /d

EW

σ δ (d

0.06
0.04
0.02

0.02 0.04 0.06 0.08

= 0

t

Y = 1

t

Y = 2

t

Y = 4

t

Y top mass uncert

CMS

Preliminary Simulation

(HATHOR inputs)

SLIDE 4

Analysis overview

Yi-Ting Duh

4

Reconstruct tt events separately with two

algorithms: – At least 4 jets – 3 jets (if one of the jets is below pT or η event selection); developed for threshold regions e l e c t r

n

/ m u

n

pTmiss

b-tagged jet b

t

a g g e d j e t jet j e t

Events / 20

20 40 60 80 100 120 140

3

10 ×

Data right reco t t wrong reco t t not reco t t bck t t single top V+jets QCD Total unc. (13 TeV)

1

35.8 fb

+jets, 3 jets µ e/

CMS

Preliminary

) t (t

t

p

50 100 150 200 250 300 350 400 450 500

Pred. Data

0.6 0.8 1 1.2 1.4

Events / 20

20 40 60 80 100

3

10 ×

Data right reco t t wrong reco t t not reco t t bck t t single top V+jets QCD Total unc. (13 TeV)

1

35.8 fb

+jets, 4 jets µ e/

CMS

Preliminary

) t (t

t

p

50 100 150 200 250 300 350 400 450 500

Pred. Data

0.6 0.8 1 1.2 1.4

Events / 20

5000 10000 15000 20000 25000 30000 35000 40000 45000

Data right reco t t wrong reco t t not reco t t bck t t single top V+jets QCD Total unc. (13 TeV)

1

35.8 fb

5 jets ≥ +jets, µ e/

CMS

Preliminary

) t (t

t

p

50 100 150 200 250 300 350 400 450 500

Pred. Data

0.6 0.8 1 1.2 1.4

Single top (Simulation) Vector+jets (Simulation) Multi-jets (Data sideband)

Lepton+jets event signature:

– One lepton (electron or muon) – Four or more jets and two are b-tagged – Missing transverse momentum

SLIDE 5

Extracting Yukawa coupling

Yi-Ting Duh

5

Likelihood fit in (Mtt, Δytt)

to constrain Yt

Events

10000 20000 30000

Data t t single top V+jets QCD

CMS

Preliminary

(13 TeV)

1

35.8 fb

3 jets ≥ +jets, µ e/

0-0.6 0.6-1.2 >1.2 3 jets 0-0.6 0.6-1.2 >1.2 4 jets 0-0.6 0.6-1.2 >1.2 5 jets ≥

]

2

[GeV/c

t t

m

0-300 300-320 320-340 340-360 360-400 400-440 440-500 500-2000 0-340 340-380 380-420 420-460 460-520 520-2000 0-400 400-440 440-480 480-520 520-580 580-660 660-2000 0-360 360-400 400-440 440-480 480-540 540-2000 0-400 400-440 440-480 480-520 520-580 580-2000 0-500 500-560 560-620 620-700 700-2000 0-360 360-400 400-440 440-480 480-540 540-2000 0-400 400-440 440-480 480-520 520-580 580-2000 0-500 500-560 560-620 620-700 700-2000

Data/Pred.

0.98 1 1.02

After the likelihood fit

t

Y

0.5 1 1.5 2 2.5 3 3.5 4

yields)/(Powheg yields)

t

Y (

0.9 0.95 1 1.05 1.1 1.15 1.2 1.25 1.3

3 jets CMS

Preliminary Simulation

bin 1

t

Y

0.5 1 1.5 2 2.5 3 3.5 4

yields)/(Powheg yields)

t

Y (

0.9 0.95 1 1.05 1.1 1.15 1.2 1.25 1.3

3 jets CMS

Preliminary Simulation

bin 2

t

Y

0.5 1 1.5 2 2.5 3 3.5 4

yields)/(Powheg yields)

t

Y (

0.9 0.95 1 1.05 1.1 1.15 1.2 1.25 1.3

3 jets CMS

Preliminary Simulation

bin 3

Example bins, more for each bin

y-axis: (At detector level) Strength of the EW correction Uncorrected MC yields

SLIDE 6

t

Y

0.5 1 1.5 2 2.5 3

ln L Δ

2

1 2 3 4 5 6

Expected Observed

CMS

Preliminary

(13 TeV)

1

35.8 fb +jets, all jets combined µ e/

68% CL Z 95% CL Z

Results

Combined fit extract to an upper limit on the top

quark Yukawa coupling of 1.67 (1.62 expected) at 95% confidence level Yi-Ting Duh

6

t

Y

0.5 1 1.5 2 2.5 3

ln L Δ

2

1 2 3 4 5 6

Expected Observed

CMS

Preliminary

(13 TeV)

1

35.8 fb

+jets, 4 jets µ e/

68% CL Z 95% CL Z t

Y

0.5 1 1.5 2 2.5 3

ln L Δ

2

1 2 3 4 5 6

Expected Observed

CMS

Preliminary

(13 TeV)

1

35.8 fb

5 jets ≥ +jets, µ e/

68% CL Z 95% CL Z t

Y

0.5 1 1.5 2 2.5 3

ln L Δ

2

1 2 3 4 5 6

Expected Observed

CMS

Preliminary

(13 TeV)

1

35.8 fb

+jets, 3 jets µ e/

68% CL Z 95% CL Z

Promising sensitivities compare to
ther measurements

tttt production constraints the Yukawa coupling indirectly <2.1 with the same dataset ≧

SLIDE 7

Backup

SLIDE 8

Control Plots, 3 jets

8

SLIDE 9

Control Plots, 4 jets

9

SLIDE 10

Control Plots, ≧5 jets

10

SLIDE 11

Uncertainties: experimental

11

All POG recommendations: ID/trigger, tagging, JEC/JER, pileup, lumi, MET variation

is negligible

JEC split into 19 independent sources → the dominant experimental uncertainty
Corresponding normalization uncertainties for backgrounds
QCD shape uncertainty derived by b-tagging inversion → larger and higher impact

uncertainties for 3 jets channel

SLIDE 12

Uncertainties: modelling

12

Renormalization & Factorization scale uncertainties are evaluated by varying each

scale independently by a factor of 2.

Top mass systematic derived by ±1GeV MC samples
Parton shower (prescribed by TOP PAG): 6 nuisances, UE tuning variation is negligible
Uncertainty due to weak correction estimated by bin-by-bin (scale variation)x(weak

correction) → tiny systematic variation and low impact

SLIDE 13

Background Estimation

13

Backgrounds of tt production decays into lepton+jets including:
Single top
V+jets: W+jets and Drell Yan+jets
QCD Multijets

Simulation Simulation Data-driven

QCD shape is obtained from data sideband defined by the maximum CSVv2 discriminator <0.6 QCD normalization is estimated by the ratio of the yields between data and MC in signal region and in sideband region

subtract contribution from V+jets, t, ttbar

SLIDE 14

QCD Systematic Uncertainty

Evaluate the shape systematics by considering the difference in shape

between regions III and IV

Normalization systematics 30% is assigned

14

signal region invert iso

(remove trigger, pT(μ)>50 GeV)

iso medium CSV CSV<0.6

Ⅰ Ⅱ Ⅳ Ⅲ

QCD shape templates

derive shape sys

SLIDE 15

Reconstruct tt events : 3 jets

Yi-Ting Duh

15

At the threshold of tt production, quarks from tt decay are likely to have pT or η outside
f the selection thresholds (mostly happens in the softer pT jet from W decays)

Assuming that the two jets with the highest value of b-tagging discriminator are associated with b-quarks, the ambiguity is the assignment of the b jets Construct a likelihood discriminant to identify the best jet-to-parton assignment

80% correctly identify

Hypothesized m(bh+Wh jets)

Solution distance

[Nucl. Instrum. Meth. A 736 (2014) 169] Select p𝑤 as the point on the ellipse for which the distance bwt pTmiss is minimal Intersect an ellipse in 3D momentum space