Precision Higgs physics: a gateway to New Physics Jonas M. Lindert - - PowerPoint PPT Presentation

Precision Higgs physics: 
 a gateway to New Physics

Jonas M. Lindert SM@LHC 2018 Higgs-session
 Berlin, 10. April 2018

Jonas M. Lindert

2

Finding the Higgs was “easy”… …finding new physics might 
 be very tough.

[CMS diphoton Higgs search, arXiv:1407.0558] [Grazzini et. al., 2016]

• Bump hunting: little to no theoretical

input needed. Look for BSM effects in small deviations from SM predictions: → Higgs processes natural place to look at → good control on theory necessary!

Jonas M. Lindert

SM ~ v.e.v. ΛNP direct bounds ~ TeV

Imagine to have new physics at a (heavish) scale ΛNP Typical modification to observable w.r.t. standard model prediction: δO ~ Q2/ΛNP 2 To gain over direct bounds:

IN THE BULK: Q~MH → few percent IN THE TAIL: Q≳ 500 GEV → ~10-20%

3

“good control” ?

[F. Caola, Moriond ’17]

Jonas M. Lindert

4

H

Outline

• ttH: ttbb background modelling
• H-inc: mixed QCD-EW
• H-pT: NNLO+N3LL
• H-pT: t & b NLO mass effects

∆TH~5% ∆TH~1-2% ∆TH~0.5% ∆TH~5%

• VBF-H: differential NNLO revised

New! New! New! New! New!

• VH(+jet) @ NLOPS QCD+EW New!

Jonas M. Lindert

5

H

Outline

• H-inc: mixed QCD-EW
• H-pT: NNLO+N3LL
• H-pT: t & b NLO mass effects

∆TH~5% ∆TH~1-2% ∆TH~0.5% ∆TH~5%

New! New! New!

• VBF-H: differential NNLO revised New!
• ttH: ttbb background modelling New!
• VH(+jet) @ NLOPS QCD+EW New!

Jonas M. Lindert

6

inclusive-H

48.58 pb = 16.00 pb (+32.9%) (LO, rEFT) + 20.84 pb (+42.9%) (NLO, rEFT) − 2.05 pb (−4.2%) ((t, b, c), exact NLO) + 9.56 pb (+19.7%) (NNLO, rEFT) + 0.34 pb (+0.2%) (NNLO, 1/mt) + 2.40 pb (+4.9%) (EW, QCD-EW) + 1.49 pb (+3.1%) (N3LO, rEFT) σ = 48.58 pb+2.22 pb (+4.56%)

−3.27 pb (−6.72%) (theory) ± 1.56 pb (3.20%) (PDF+αs) .

δ(scale) δ(trunc) δ(PDF-TH) δ(EW) δ(t, b, c) δ(1/mt)

+0.10 pb −1.15 pb

±0.18 pb ±0.56 pb ±0.49 pb ±0.40 pb ±0.49 pb

+0.21% −2.37%

±0.37% ±1.16% ±1% ±0.83% ±1%

13 TeV:

• remaining uncertainties are at the ~1%-level (and basically everything becomes relevant):

αS2 αS3 αS4 αS5

[Anastasiou, Melnikov; Harlander, Kilgore] [Anastasiou et al]

[Anastasiou et al.;2015] [Harlander, Kilgore; 2002 Anastasiou, Melnikov;2002] [Dawson; 1991 Djouadi, Spira, Zerwas;1991]

Obtained through a series expansion around the soft limit

1 − z = 1 − m2

H/ˆ

s

[Anastasiou et al.;2016]

Jonas M. Lindert

7

inclusive-H

48.58 pb = 16.00 pb (+32.9%) (LO, rEFT) + 20.84 pb (+42.9%) (NLO, rEFT) − 2.05 pb (−4.2%) ((t, b, c), exact NLO) + 9.56 pb (+19.7%) (NNLO, rEFT) + 0.34 pb (+0.2%) (NNLO, 1/mt) + 2.40 pb (+4.9%) (EW, QCD-EW) + 1.49 pb (+3.1%) (N3LO, rEFT) σ = 48.58 pb+2.22 pb (+4.56%)

−3.27 pb (−6.72%) (theory) ± 1.56 pb (3.20%) (PDF+αs) .

δ(scale) δ(trunc) δ(PDF-TH) δ(EW) δ(t, b, c) δ(1/mt)

+0.10 pb −1.15 pb

±0.18 pb ±0.56 pb ±0.49 pb ±0.40 pb ±0.49 pb

+0.21% −2.37%

±0.37% ±1.16% ±1% ±0.83% ±1%

13 TeV:

• remaining uncertainties are at the ~1%-level (and basically everything becomes relevant):

αS2 αS3 αS4 αS5

[Anastasiou, Melnikov; Harlander, Kilgore] [Anastasiou et al]

[Harlander, Kilgore; 2002 Anastasiou, Melnikov;2002] [Dawson; 1991 Djouadi, Spira, Zerwas;1991]

Obtained through a series expansion around the soft limit

1 − z = 1 − m2

H/ˆ

s

[Anastasiou et al.;2015
 Mistlberger; 2018]

exact!

[Anastasiou et al.;2016]

Jonas M. Lindert

8

inclusive-H

48.58 pb = 16.00 pb (+32.9%) (LO, rEFT) + 20.84 pb (+42.9%) (NLO, rEFT) − 2.05 pb (−4.2%) ((t, b, c), exact NLO) + 9.56 pb (+19.7%) (NNLO, rEFT) + 0.34 pb (+0.2%) (NNLO, 1/mt) + 2.40 pb (+4.9%) (EW, QCD-EW) + 1.49 pb (+3.1%) (N3LO, rEFT) σ = 48.58 pb+2.22 pb (+4.56%)

−3.27 pb (−6.72%) (theory) ± 1.56 pb (3.20%) (PDF+αs) .

δ(scale) δ(trunc) δ(PDF-TH) δ(EW) δ(t, b, c) δ(1/mt)

+0.10 pb −1.15 pb

±0.18 pb ±0.56 pb ±0.49 pb ±0.40 pb ±0.49 pb

+0.21% −2.37%

±0.37% ±1.16% ±1% ±0.83% ±1%

13 TeV:

• remaining uncertainties are at the ~1%-level (and basically everything becomes relevant):

αS2 αS3 αS4 αS5

[Anastasiou, Melnikov; Harlander, Kilgore] [Anastasiou et al]

[Harlander, Kilgore; 2002 Anastasiou, Melnikov;2002] [Dawson; 1991 Djouadi, Spira, Zerwas;1991]

Obtained through a series expansion around the soft limit

1 − z = 1 − m2

H/ˆ

s

exact!

[Anastasiou et al.;2015
 Mistlberger; 2018] [Anastasiou et al.;2016]

Jonas M. Lindert

9

inclusive-H: mixed QCD-EW @ NLO

[Bonetti, Melnikov, Tancredi; ‘17+’18] q g V=W,Z H g g

• requires very complicated three-loop virtuals and two-loop reals.
• literature result [Anastasiou, Boughezal, Petriello; ’09] based on 


unphysical limit MV ≫ MH yields 5(±1)% contribution.

• this limit corresponds to a point-like ggH QCD-EW coupling.
• Two-loop reals not yet available, but (improved) soft-gluon 


approximation known to be quite reliable for inclusive Higgs.

• Only non-universal ingredient required for soft gluon approximaton: 


three-loop virtuals


• Result:
• Further improvements and updated uncertainty estimate requires

computation of the very challenging two-loop reals!

q g V H g g σLO

QCD = 20.6 pb,

σLO

QCD/EW = 21.7 pb,

σNLO

QCD = 37.0 pb,

σNLO

QCD/EW = 39.0 pb. }

consistent!

σ(N)LO

QCD/EW/σ(N)LO QCD

= 5.3 − 5.4%

?????

Jonas M. Lindert

10

Higgs-pT / Higgs+jet

X H

?

• Motivation:
• Higgs-pT sensitive probe of New Physics 


➜ In particular: disentangle cg vs. ct

• Possibility to constrain the charm-Yukawa coupling


[Bishara, Haisch, Monni, Re; ’16]

• When we are inclusive in the radiation recoiling

against the Higgs we can measure its pT.

• At LO in inclusive-H: pTH = 0 


➜ Consider Higgs+jet production

• Two regimes:
• [Grazzini et. al., 2016]

pT ⌧ mH

pT ≥ mH

• fixed-order reliable
• ggH NOT point-like
• resummation mandatory
• ggH point-like
• bottom mass effects relevent

+∞ ∞

• [Plot: Grazzini]

~ ct ~ cg

~ cg

Jonas M. Lindert

11

Higgs-pT: higher-order corrections

integrate-out

NNLO

Bottleneck: IR subtraction

H g H g H

NLO

Bottleneck: 
 massive two-loop amplitudes Idea: QCD corrections factorize 
 ➜ apply K-factors from HEFT to lower

• rder predictions in full theory ➜ check!!

NNLO+NXLL

[Bonciani et. al., ’16]

• [Chen et.al.; ’14+‘16

Boughezal et. al.; ’15,
 Caola et.al.; ’15]

• perturb. uncertainties in HEFT

under very good control:

• ~10% scale variation
• stable shapes

[Chen et.al.; ’16] pTH

Ansätze:

• analytical: very hard, planar MI known
• numerical: very CPU/GPU intensive
• expansions: has to be performed carefully, very versatile

[Jones et. al., ’18] [Melnikov et. al., ’16+’17] full theory: loop-induced HEFT: tree-level at LO heavy quarks

Jonas M. Lindert

ratio to N3LL pt

H [GeV]

0.7 0.8 0.9 1 1.1 1.2 1.3 5 10 15 20 25 30 35 40

12

: N3LL+NNLO

pH

T ⌧ mH [Bizon, Monni, Re, Rottoli, Torrielli; 2017]

RadISH, 13 TeV, mH = 125 GeV µR = µF = mH, Q = mH/2 PDF4LHC15 (NNLO) uncertainties with µR, µF, Q variations (x 3/2) Fixed order from PRL 115 (2015) 082003

1/σ dσ/d pt

H [1/GeV]

pt

H [GeV]

NNLO NNLL+NNLO N3LL+NNLO 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 10 20 30 40 50 60 70 80 90 100 110 120

NNLL/N3LL

• almost all ingredients for consistent N3LL

resummation are available (apart from four- loop cusp anomalous dimension) 
 [Li, Zhu; ’16] & [A.Vladimirov; 16]

• application to Higgs-pT is based on a new

formulation of momentum-space resummation

• matched to NNLO [Caola et.al.; ’16]
• NNLL/N3LL:
• ~10-20% effects around Sudakov peak
• reduced scale dependence at N3LL
• NNLL+NNLO / N3LL+NNLO:
• few percent around Sudakov peak
• several percent below Sudakov peak
• scale dependence very similar


➡ Resummation effects under very good control! matched to NNLO (multiplicative): pure logarithmic:

Jonas M. Lindert

ratio to N3LL pt

H [GeV]

0.7 0.8 0.9 1 1.1 1.2 1.3 5 10 15 20 25 30 35 40

13

: N3LL+NNLO

pH

T ⌧ mH [Bizon, Monni, Re, Rottoli, Torrielli; 2017]

RadISH, 13 TeV, mH = 125 GeV µR = µF = mH, Q = mH/2 PDF4LHC15 (NNLO) uncertainties with µR, µF, Q variations (x 3/2) Fixed order from PRL 115 (2015) 082003

1/σ dσ/d pt

H [1/GeV]

pt

H [GeV]

NNLO NNLL+NNLO N3LL+NNLO 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 10 20 30 40 50 60 70 80 90 100 110 120

NNLL/N3LL

• almost all ingredients for consistent N3LL

resummation are available (apart from four- loop cusp anomalous dimension) 
 [Li, Zhu; ’16] & [A.Vladimirov; 16]

• application to Higgs-pT is based on a new

formulation of momentum-space resummation

• matched to NNLO [Caola et.al.; ’16]
• NNLL/N3LL:
• ~10-20% effects around Sudakov peak
• reduced scale dependence at N3LL
• NNLL+NNLO / N3LL+NNLO:
• few percent around Sudakov peak
• several percent below Sudakov peak
• scale dependence very similar


➡ Resummation effects under very good control! matched to NNLO (multiplicative): pure logarithmic:

50 100 150 200 250 300 350

2 −

10

1 −

10 1 [fb/GeV]

γ γ T

p / d

fid

σ d ATLAS

• 1

= 13 TeV, 36.1 fb s

,

γ γ → H

Data, tot. unc.

• Syst. unc.

XH default MC + H → gg

bbH + ttH + VH = VBF+ XH

XH HRes 2.3 + XH +

RadISH+NNLOJET

50 100 150 200 250 300 350 [GeV]

γ γ T

p 1 2

XH Ratio to default MC +

CERN-EP-2017-288

Jonas M. Lindert

14

top & bottom mass effects

LO

pT ⌧ mH:

pT ≥ mH:

• top-loop gets resolved by hard radiation
• point-like ggH and full theory have very

different high energy behaviour:

ng

∼ p−2

T

∼ p−4

T

[Caola, F

HEFT full theory

➡What about NLO corrections? 
 They are quite sizeable in the HEFT….

[Grazzini, Sargsyan; ’13]

• not just just top-loop contributes:


• ➡ Sudakov-like logarithmic enhancement of 


b-quark contribution at small pT

m2

b/m2 H log2(p2 ⊥/m2 b) [

not accounted for in

d Ab

gg→Hg ⇠

~10% ~1%

pT ⌧ mH

for

dσ/dpT ∝ y2

t + ytyb + y2 b

Jonas M. Lindert

15

: bottom mass effects at NLO

pH

T ⌧ mH [JML, Melnikov, Tancredi, Wever ’17]

• -(5-10)% for pT=20-40 GeV at LO and NLO
• Despite (large) corrections, the interference

shape stable under QCD corrections 
 → solid observable

• large mb-renormalisation scheme

dependence tamed at NLO Outlook: combine with NNLO+N3LL

• expansion of the two-loop integrals in 


[Melnikov, Tancredi, Wever; ‘16+‘17]

• valid at %-level down to pT~10 GeV
• real radiation treated exact with OpenLoops

(m2

b/p2 T)

Jonas M. Lindert

16

: top mass effects at NLO

pH

T ≥ mH Control of the high-H-pT tail at NLO opens the door for new physics searches in this regime!

10−6 10−5 10−4 10−3 10−2 10−1 100 101 102 dσ/dpT,H [fb/GeV]

pp → H + j @ 13 TeV LOHEFT NLOHEFT LO NLO(m0

H,m2 t)

400 600 800 1000 1200 1400 1600 1800 p⊥ [GeV] 1.0 1.5 2.0 2.5 NLO/LO

10−7 10−6 10−5 10−4 10−3 10−2 10−1 100 dσ/dpt, H [pb/GeV] ratio NLO/LO

LHC 13 TeV PDF4LHC15 NLO µ = HT

2

10−1 100

ratio to LO HEFT

1.0 2.0 200 400 600 800 1000 LO HEFT NLO HEFT LO Full NLO Full 10−1 100 pt, H [GeV] 1.0 2.0 200 400 600 800 1000

confirmed at NLO nea

[JML, Kudashkin, Melnikov, Wever; ’18]

• NLO corrections very similar as in HEFT: K~2 with remaining scale uncertainties ~20-25%
• hardly any shape dependence
• numerical integration of two-loop integrals

based on SecDec [Borowka et.al.]

• valid in all of the phase-space
• expansion of the two-loop integrals up 


to at the level of the DE

[Kudashkin, Melnikov, Wever; ‘17]

• valid at %-level for large pT

[Jones, Kerner, Luisoni; ’18]

(m2

t/p2 T)1, (m2 H/p2 T)0

Jonas M. Lindert

17

H

Outline

∆TH~5% ∆TH~1-2% ∆TH~0.5% ∆TH~5%

• ttH: ttbb background modelling New!
• VH(+jet) @ NLOPS QCD+EW New!
• H-inc: mixed QCD-EW
• H-pT: NNLO+N3LL
• H-pT: t & b NLO mass effects

New! New! New!

• VBF-H: differential NNLO revised New!

Jonas M. Lindert

18

VBF-H: differential NNLO

• In particular for VBF the inclusive cross section can not be measured: 


tight VBF cuts on tagging jets (large mj1j2 & Δyj1j2) have to be imposed ➡Fully differential predictions mandatory

• Furthermore: pTj1/2 proxy for -q2 ➜ sensitive to NP
• 2→3 NNLO possible in DIS/VBF approximation (neglecting cross-talk)
• first differential computation: [Cacciari, Dreyer, Karlberg, Salam, Zanderighi; ’15]
• Sizeable differential NNLO/NLO corrections:

10-15% in the tails

• …not covered by NLO scale variations

Jonas M. Lindert

19

VBF-H: differential NNLO

• In particular for VBF the inclusive cross section can not be measured: 


tight VBF cuts on tagging jets (large mj1j2 & Δyj1j2) have to be imposed ➡Fully differential predictions mandatory

• Furthermore: pTj1/2 proxy for -q2 ➜ sensitive to NP
• 2→3 NNLO possible in DIS/VBF approximation (neglecting cross-talk)
• independent differential computation: [Cruz-Martinez, Gehrmann,Glover, Huss ’18]
• Still relevant NNLO/NLO corrections,

but smaller: ~5% in the tails

• Corrections mostly covered by scale

variations


• Origin of difference: error in previous
• results. Now fixed.

10-1 100 101 NNLOJET

√s ‾ = 13 TeV

dσ/dpT

j1[fb/GeV]

VBF H 2j NNLO

LO NLO NNLO 0.8 0.9 1 1.1 50 100 150 200 250 300 NNLOJET

√s ‾ = 13 TeV

Ratio to NLO pT

j1 [GeV]

10-3 10-2 10-1 100 101 NNLOJET

√s ‾ = 13 TeV

dσ/dpT

j2[fb/GeV]

VBF H 2j NNLO

LO NLO NNLO 0.8 0.9 1 1.1 50 100 150 200 250 300 NNLOJET

√s ‾ = 13 TeV

Ratio to NLO pT

j2 [GeV]

➡ These calculations are very complex: 
 cross-validation is important!

Jonas M. Lindert

20

H

Outline

∆TH~5% ∆TH~1-2% ∆TH~0.5% ∆TH~5%

• H-inc: mixed QCD-EW
• H-pT: NNLO+N3LL
• H-pT: t & b NLO mass effects

New! New! New!

• VBF-H: differential NNLO revised New!
• ttH: ttbb background modelling New!
• VH(+jet) @ NLOPS QCD+EW New!

Jonas M. Lindert

Motivation: • HV with H→bb allows to constrain Hbb coupling

• only feasible in boosted regime pT,V > 150-200 GeV
• in this regime impact of (negative) virtual EW Sudakov corrections

HV(+jet) at NLO+PS QCD+EW

[Granata, JML, Oleari, Pozzorini; ’17]

pT of Higgs

• NLO EW: -10(-20)% > 200(800) GeV
• MiNLO ensures NLO QCD and NLO EW 


accuracy in the whole phase-space

10−8 10−7 10−6 10−5

dσ/dpH

T [pb/GeV]

13 TeV HW −j Boosted regime MiNLO+PS dσ/dσQCD

0.6 0.8 1.0 200 1000

dσ/dpH

T [pb/GeV]

13 TeV HW −j Boosted regime MiNLO+PS dσ/dσQCD QCD+EW QCD pH

T [GeV]

0.6 0.8 1.0 200 1000

, pV

T ≥ 190 GeV .

:

POWHEG-BOX+OpenLoops 10−6 10−5 10−4 10−3

dσ/dMe+e− [pb/GeV] 13 TeV HZ NLO+PS dσ/dσQCD

0.5 1.0 60 70 80 90 100 110 120 130 140

dσ/dMe+e− [pb/GeV] 13 TeV HZ NLO+PS dσ/dσQCD QCD QCD (no QED shower) QCD+EW Me+e− [GeV]

0.5 1.0 60 70 80 90 100 110 120 130 140 POWHEG-BOX+OpenLoops

Mll

• large QED effects due to radiative tail…
• …reliably modelled by QED-PS
• matching at NLO EW has to be resonance-aware

(as provided by POWHEG-BOX-RES) [Jezo & Nason; ’15]

Jonas M. Lindert

22

H

Outline

• ttH: ttbb background modelling

∆TH~5% ∆TH~1-2% ∆TH~0.5% ∆TH~5%

• VH(+jet) @ NLOPS QCD+EW New!
• H-inc: mixed QCD-EW
• H-pT: NNLO+N3LL
• H-pT: t & b NLO mass effects New!
• VBF-H: differential NNLO revised New!

Jonas M. Lindert

23

ttH: taming the ttbb background

Breaking News

H→bb: already with observation 
 starting to be systematics limited

Jonas M. Lindert

24

ttH: taming the ttbb background

State-of-the-sart: NLOPS in 4FS (with mb>0)

• 20–30% scale uncertainties
• matching/shower effects up to 30% in Higgs region
• due to double g→bb splittings (one splitting from PS!) [Cascioli et al ’13]
• NLOPS matching & shower systematics need to be understood in detail!
• Note: very complex final state (top decays are included)
• Large differences between MC tools 


→ need to be understood!

• ongoing effort within HXSWG (comparing 8 NLOPS tools)
• goal: recommendations for tt+b-jet uncertainties

decayed ttbb parton SherpaOL MG5+PY8(.5,1)HT/2 nosc MG5+HW++(.5,1)HT/2 nosc MBx+HW7 (.7,1)HT/2 PowhegOL+PY8 PowhegOL+HW7 10−5 10−4 10−3 pT of 1st light-jet (WW4b cuts) dσ/dpT [pb/GeV] 2.5 Y 50 100 150 200 250 300 350 400 0.5 1 1.5 2 2.5 pT [GeV] ratio to PwgOL+PY

Jonas M. Lindert

25

Conclusions

• Without a clear sign of new-physics it is crucial to perform detailed

theory vs. experiment comparisons, to look for possible deviations. ➡ Higgs is an obvious place to look at!


• To this end we have to understand (and measure) inclusive Higgs

production at the percent level ➡ For remaining theory uncertainties: leave no stone unturned.

• …and differential Higgs production at the ~10% level

➡ Feasible due to advances in perturbative calculations:
 NNLO, N3LL, NLO mass effects, NLO EW, NLOPS

• Outlook: