NLO for Higgs signalshiggs/talks/heinrich.pdf · Gudrun Heinrich Max Planck Institute for Physics, Munich NLO for Higgs signals

Gudrun HeinrichMax Planck Institute for Physics, Munich

NLO for Higgs signals

P.Meridiani, EPS 2017

P.Meridiani, EPS 2017

we need to be sure about the systematic uncertainties to see something like this

status of higher (fixed) order predictions involving H

gg ! H

VBF

WH /ZH

(see also Kirill’s talk)

ttH

H + jets

H + single top

HH

�tot

d� bonus

mt ! 1black: (HEFT)

N^3LO NNLO N^3LL threshold resum.NLO EW, mixed QCD/EW

N^3LO NNLO

NNLO NNLO NLO EW

NLO EW

NLO NLO NNLL threshold resum.

NLO NLO

NLO EW

1j: NNLO 1j: NNLO

NNLO NNLO NLO

H+1jet NLO: mb dependenceNLO up to 3jets

NLL resum.

NNLL resum.

blue: dependencemq

gg to ZH mt dep. approx.

status of higher (fixed) order predictions involving H

gg ! H

VBF

WH /ZH

ttH

H + jets

H + single top

HH

�tot

d� bonusN^3LO NNLO (+NNLL) N^3LL threshold resum.

NLO EW, mixed QCD/EW

N^3LO NNLO

NNLO NNLO NLO EW

NLO EW

NLO NLO

NNLL threshold resum.NLO NLO

NLO EW

1j: NNLO 1j: NNLO

NNLO NNLONLO

H+1jet NLO: mb dependence

NLO up to 3jets

Beenakker et al ’01; Dawson, Reina ’02; Frederix et al ‘14

Borowka, Greiner, GH, Jones, Kerner, Schlenk, Schubert, Zirke ‘16

Lindert, Melnikov, Tancredi, Wever ‘17

Frixione et al ’14; (on-shell)Denner, Lang, Pellen, Uccirati ’16

Kulesza et al ‘15

Anastasiou et al ’13-’15

Cullen et al ’13

Boughezal, Caola et al ’15; Boughezal et al ’15; Chen et al ’15; Dulat, Mistlberger ‘17

De Florian, Mazzitelli ’15; Steinhauser et al ‘15 De Florian et al ’16

Demartin, Maltoni, Mawatari, Zaro ’15

Brein, Djouadi, Harlander ’03;Brein, Harlander, Wiesemann, Zirke ’11

Ferrera, Grazzini, Tramontano ’11,’14

Denner, Dittmaier, Kallweit, Mück ‘12gg to ZH 1/mtAltenkamp et al ‘12

NLL resum. Ferrera, Pires ’16

NNLL threshold resum.

Ciccolini, Denner, Dittmaier ‘07Bolzoni, Maltoni, Moch, Zaro ’10;

Dreyer, Karlberg ‘16Cacciari, Dreyer, Karlberg,

Salam, Zanderighi ‘15

Dulat, Mistlberger ’17

Bizon, Monni, Re, Rottoli, Torrielli ‘17

Harlander et al ’14

De Florian, Grazzini, Tommasini ’11,‘12Degrassi, Maltoni ’04; Actis, Passarino, Sturm, Uccirati ’08; Anastasiou et al ‘08

status of parton shower matched results

gg ! H

VBF

ttH

H + jets

H + single top

HH

NNLO

NLO

NLO

NLO

NLO

WH /ZH

Frederix, Frixione, Vryonidou, Wiesemann ‘16

GH, Jones, Kerner, Luisoni, Vryonidou ’17; Jones, Kuttimalai ‘17

NLO up to 2jetsBuschmann, Goncalves, Krauss, Kuttimalai, Schönherr, Plehn ‘14

Frederix et al ’11; Garzelli et al ’11; Hartanto et al ’15; QCD and EW: Denner, Lang, Pellen, Uccirati ‘16

NLO (+ 0/1-jet merging) Luisoni, Nason, Oleari, Tramontano ’13

NLO QCD+EW Granata, Lindert, Oleari, Pozzorini ’17

Goncalves, Krauss, Kuttimalai, Maierhöfer ’15; Astill, Bizon, Re, Zanderighi’ 16

NNLO

Demartin, Maltoni, Mawatari, Zaro ’15

Hamilton, Nason, Re, Zanderighi ’13; Höche, Li, Prestel ’14;Alioli et al ’13,‘15

Nason, Oleari ’10; Frixione, Torrielli, Zaro ’13; Campanario et al ’13; Jäger et al ‘14

Higgs@NLO

some pressing problems to solve at NLO:

• quark mass effects “boosted Higgs can resolve loops from heavy BSM particles” and heavy SM particles (top)!

• EW corrections, combination QCD/EW

• multi-jet merging

• improve logarithmic accuracy of parton showers

• finding “optimal” scales/assessment of scale uncertainties

• consistent combination of EFT and SM NLO corrections

(also shower/resummation scale uncertainties)

Higgs pT

Forte, Muselli ’15;Marzani, Ball, del Duca, Forte, Vicini ’08; high-energy resummation of Higgs pT in gg ! H with off-shell gluons

d�/dp2T,h !�p2T,h

��1

for pT,h ! 1 :

d�/dp2T,h !�p2T,h

��2

Caola, Forte, Marzani, Muselli, Vita ’16


quark mass effects in Higgs + jetsGreiner, Höche, Luisoni, Schönherr, Winter ’16

cross sections (+Ntuples) for H+1,2,3 jets calculated at

• NLO in HEFT ( mt ! 1 limit )• LO with full mt,mb dependence

default scale choice:

basic cuts:

quark mass effects in Higgs + jets

• different jet multiplicities show very similar behaviour

mt ! 1 limit starts to fail at pT,H ⇠ 200GeV

• leading jet pT distribution also similar

suggests that resolution of top quark loops is driven by pT,Hor largest pT in the event

jet multiplicity seems to play minor role

needs further investigation (backup slide)

Effect of VBF cuts

VBF cuts can enhance the deviations of HEFT to full theory

bottom quark effects

• well below the scale uncertainties even at low pT

• depend on jet multiplicity

• destructive interference between top- and bottom-quark loops for

H+1jet

Frederix, Frixione, Vryonidou, Wiesemann ‘16

Higgs+0,1,2 jets + parton shower

see also Buschmann, Goncalves, Krauss, Kuttimalai, Schönherr, Plehn ‘14

H + n jets:

• exact 2-loop virtual for n=0

• rescaled HEFT virtual for n=1,2

• exact mass dependence in real radiation

Higgs + jetLindert, Melnikov, Tancredi, Wever ‘17

Chris Wever, RadCor 2017

Higgs + jetLindert, Melnikov, Tancredi, Wever ‘17

• large relative corrections top-bottom interference to top-top

• large mb renormalisation ambiguities (light bands)reduced at NLO, in particular at low pT

Higgs + jet

Neumann, Williams ‘17• exact mass dependence

in real radiation

1/mt expansion • virtual part:

NLO*:

on amplitude level

fullasy: expansion on amplitude squared level

• method works well for scales up to about 300 GeV

towards Higgs + jet at full NLOJones, Kerner, Luisoni

Frederix, Frixione, Hirschi, Maltoni, Mattelaer, Torrielli, Vryonidou, Zaro ‘14

�ggHH ⇠ 10�3 �ggH

largest cross section from gluon fusion, but still

Higgs boson pair production

LO with full heavy quark mass dependence Glover, van der Bij ’88, Plehn, Spira, Zerwas ’96

Higgs boson pair production in gluon fusion

Note: limit: “Higgs Effective Field Theory” (HEFT)mt ! 1

HH production threshold: 2mH <ps

HEFT strictly valid only for ps ⌧ 2mt validity of HEFT limited to

250GeV <ps < 340GeV

)o

LO with full heavy quark mass dependence Glover, van der Bij ’88, Plehn, Spira, Zerwas ’96


Note: limit: “Higgs Effective Field Theory” (HEFT)mt ! 1

HH production threshold: 2mH <ps

HEFT strictly valid only for ps ⌧ 2mt validity of HEFT limited to

250GeV <ps < 340GeV

)o

NLO in Born-improved HEFT Dawson, Dittmaier, Spira ’98 (HPAIR)

Frederix, Hirschi, Mattelaer, Maltoni, Torrielli, Vryonidou, Zaro ’14; Maltoni, Vryonidou, Zaro ’14

-10%• full mass dependence in NLO real radiation (“FTapprox”)

Grigo, Hoff, Melnikov, Steinhauser ’13, ’15 ; Degrassi, Giardino, Gröber ’16 (±10%)• supplemented with expansion: 1/mt

“Born-improved NLO HEFT”: rescale by MLO(mt)/MLOHEFT

K ' 2

• soft gluon resummation NNLL Shao, Li, Li, Wang ’13; De Florian, Mazzitelli ‘15

NNLO in mt ! 1 limit:

De Florian, Mazzitelli ’13

• including all matching coefficients Grigo, Melnikov, Steinhauser ’14

+20%

Grigo, Hoff, Steinhauser ’15 • supplemented with expansion: 1/mt


• differential NNLO De Florian, Grazzini, Hanga, Kallweit, Lindert, Maierhöfer, Mazzitelli, Rathlev ‘16

• total xs NNLO

+9%

• soft gluon resummation NNLL Shao, Li, Li, Wang ’13; De Florian, Mazzitelli ‘15

NNLO in mt ! 1 limit:

De Florian, Mazzitelli ’13

• including all matching coefficients Grigo, Melnikov, Steinhauser ’14

+20%

Grigo, Hoff, Steinhauser ’15 • supplemented with expansion: 1/mt


• differential NNLO De Florian, Grazzini, Hanga, Kallweit, Lindert, Maierhöfer, Mazzitelli, Rathlev ‘16

• total xs NNLO

+9%

NLO calculation with full top mass dependenceBorowka, Greiner, GH, Jones, Kerner, Schlenk, Schubert, Zirke ‘16

g

g

t

H

H

4 independent scales s12, s23, mH, mtall integrals calculated numerically with

SecDecBorowka, GH, Jones, Kerner, Schlenk, Zirke ‘15Borowka, GH, Jahn, Jones, Kerner, Schlenk, Zirke ‘17

Ferrera, Pires ‘16• resummation NLL+NLOqT

graphics by S.Jones

S.Borowka, GH, S.Jones, M.Kerner, J.Schlenk, T.Zirke ‘15

version 3.0:

algorithm:

http://secdec.hepforge.org

T. Binoth, GH ‘00version 1.0:version 2.0: S.Borowka, J. Carter, GH ‘12

J. Carter, GH ‘10

pySecDec: S.Borowka, GH, S.Jahn, S.Jones, M.Kerner, J.Schlenk, T.Zirke ‘17

numerical evaluation of multi-loop integrals https://github.com/mppmu/secdec/releases


https://github.com/mppmu/secdec/releases

S.Borowka, GH, S.Jones, M.Kerner, J.Schlenk, T.Zirke ‘15

version 3.0:

algorithm:


T. Binoth, GH ‘00version 1.0:version 2.0: S.Borowka, J. Carter, GH ‘12

J. Carter, GH ‘10

pySecDec: S.Borowka, GH, S.Jahn, S.Jones, M.Kerner, J.Schlenk, T.Zirke ‘17

numerical evaluation of multi-loop integrals

can be used as an integral library

New!




• integrals calculated numerically with SecDec

• total number of integrals: g

g

t

H

H

calculation: building blocks

• amplitude reduction with Reduze [C. Studerus, A. v.Manteuffel]

• non-planar integrals computed mostly without reduction

• real radiation: (a) GoSam-1L + Catani-Seymour

dipole subtraction(b) GoSam-1L + POWHEG

• amplitude generation with 2 setups (custom made and “GoSam-2loop”)

• before reduction: ~10000, after reduction ~330, after sector decomposition 11244 (3086 non-planar)

• used finite basis for planar integrals

top mass effectstotal cross sections at 14 TeV

HXSWG:

• relative difference Born-improved NLO HEFT to full NLO:14 TeV: 16.4%

100 TeV: 31.5%27 TeV: 21.2%

scale uncertainties

top mass effects: energy dependence

preliminary, ± 0.3 stat. uncertainty

Higgs boson pair invariant mass

0

1

2

3

4

5

6

d�/d

mhh[fb/G

eV]

LO

B-i. NLO HEFT

NLO FTapprox

LO basic HEFT

NLO basic HEFT

NLO

300 400 500 600 700 800 900 1000mhh [GeV]

0.51.01.52.0

Kfa

ctor

0.00

0.05

0.10

0.15

0.20

d�/d

mhh[fb/G

eV]

LO

B-i. NLO HEFT

NLO FTapprox

LO basic HEFT

NLO basic HEFT

NLO

300 400 500 600 700 800 900 1000mhh [GeV]

0.51.01.52.0

Kfa

ctor

14 TeV 100 TeV

Born-improved NLO HEFT overestimates by about 50%, FTapprox by about 40% (at 14 TeV, worse at 100 TeV)

for large invariant masses:

top quark loops resolved HEFT has wrong scaling behaviour at high energies

rapidity of the Higgs boson pair

14 TeV 100 TeV

0.00

0.05

0.10

0.15

0.20

0.25

d�/d

mhh[fb/G

eV]

LO

NLO

NLO-i. NNLO HEFT

300 400 500 600 700 800 900 1000mhh [GeV]

0.51.01.52.0

Kfa

ctor

bin-by-bin rescaling at observable level by NNLO HEFT K-factor

d�NLO�i.NNLOHEFT

dmhh=

d�NLO

dmhh⇥ d�HEFT

NNLO/dmhh

d�HEFTNLO /dmhh

NLO-improved NNLO HEFTNNLO HEFT:

De Florian, Grazzini, Hanga, Kallweit, Lindert, Maierhöfer, Mazzitelli, Rathlev 1606.09519

“NLO-improved NNLO HEFT”: [Borowka, Greiner, GH, Jones, Kerner, Schlenk, Zirke 1608.04798]

�0 = 38.56 fb

would lead to

variation of triple Higgs coupling

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

d�/d

mhh[fb/G

eV]

LO

B-i. NLO HEFT

NLO FTapprox

LO basic HEFT

NLO basic HEFT

NLO

300 400 500 600 700 800 900 1000mhh [GeV]

0.51.01.52.0

Kfa

ctor

� = 0

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

d�/d

mhh[fb/G

eV]

LO

B-i. NLO HEFT

NLO FTapprox

LO basic HEFT

NLO basic HEFT

NLO

300 400 500 600 700 800 900 1000mhh [GeV]

0.51.01.52.0

Kfa

ctor

� = �1

� = �BSM/�SM

�1 0 1 2 3 4 5

�

0

20

40

60

80

100

120

140

160

�[fb]

LO

NLO

NLO HEFT

NLO FTapprox

� = 2

cross section has a minimum around due to destructive interference

between diagrams containing �and box-type diagrams

0.0

0.2

0.4

0.6

0.8

1.0

1.2

d�/d

mhh[fb/G

eV]

LO

B-i. NLO HEFT

NLO FTapprox

LO basic HEFT

NLO basic HEFT

NLO

300 400 500 600 700 800 900 1000mhh [GeV]

0.51.01.52.0

Kfa

ctor

� = 5

degeneracy due to quadratic dependence�

distributions can discriminate between degenerate values�

variation of triple Higgs coupling

GH, S.Jones, M.Kerner, G.Luisoni, E.Vryonidou ‘17

• two-loop amplitude depends only on s, t (mt,mH fixed)

• variable transformation to achieve more uniform distribution

construct 2-dim grid

• avoid evaluation of two-loop amplitude for each phase space point

�(s) =q

1� 4m2H/s

combination with POWHEG and MadGraph5_aMC@NLO

and Sherpa

POWHEG-BOX-V2: User-Process-V2/ggHH

combination with parton showers

GH, S.Jones, M.Kerner, G.Luisoni, E.Vryonidou ‘17

• two-loop amplitude depends only on s, t (mt,mH fixed)

• variable transformation to achieve more uniform distribution

construct 2-dim grid

• avoid evaluation of two-loop amplitude for each phase space point

�(s) =q

1� 4m2H/s

combination with POWHEG and MadGraph5_aMC@NLO

and Sherpa New!

see Silvan Kuttimalai’s talk

POWHEG-BOX-V2: User-Process-V2/ggHH

combination with parton showers

grid validation

slide: S.Jones

hdamp = 1

hdamp=h limits amount of exponentiated hard radiation

dependence on shower parameters

hdamp = 1

hdamp=h limits amount of exponentiated hard radiation

dependence on shower parameters

shower effects large but order(s) of magnitude smaller than difference to Born-improved HEFT

• NLO is not “a solved problem”• finite mass effects are important

• in particular in tails of distributions, where they need to be distinguished from BSM effects

• combination of QCD corrections with NLO EW and EFT

pp ! HH, pp ! H + jet, gg ! HZ

means 2-loop integrals with many kinematic scales (too many for an analytic solution currently?) • HH@NLO done numerically (SecDec)

• NLO with full top mass dependence for

• increase precision by improvements on parton shower side (matching uncertainties, merging, log. acc.) and resummation

talk of Pier-Francesco Monni

talk of Alexander Mück many talks

lots of progress recently

summary & outlook

BACKUP SLIDES

compare POWHEG and MG5_aMC@NLO

MG5_aMC@NLO old shower starting scale Qsh:

MG5_aMC@NLO version 2.5.3:

picked with some probability distribution in old Qsh

new Qshnew Qshpicked with some probability distribution in

compare POWHEG and MG5_aMC@NLO

new default Qsh matches onto NLO fixed order at large pT

compare Pythia6 and Pythia8

ratio to lower jet multiplicity shows similar behaviour


quark mass effects in Higgs + jetsGionata Luisoni, DIS 2017

10�1

100

X/

LO

H+

2

no veto applied

Ratio wrt. LO using mt ! 1 approximation.

0 100 200 300 400 500 600 700 800 900 1000

Higgs boson transverse momentum: pT, H [GeV]

10�1

100

X/

LO

H+

2(v

eto)

pT, H < 100 GeV


10�8

10�7

10�6

10�5

10�4

10�3

10�2

10�1

100

101

d�/d

p T,H

[pb/G

eV]

GoSam + Sherpapp ! H + 3 jets at 13 TeV

CT14nlo, R = 0.4 anti-kT, |⌘jet| < 4.4, pT,jet = 100 GeV

LO H+3 (⇥10)

LO H+3 (veto)

LO H+3 mt,b (⇥10)

LO H+3 mt,b (veto)

LO H+3 mt (⇥10)

LO H+3 mt (veto)


10�1

100

X/

LO

H+

2no veto applied


100 200 300 400 500 600 700 800 900 1000Leading-jet transverse momentum: pT, j1 [GeV]

10�1

100

X/

LO

H+

2(v

eto)

pT, H < 100 GeV


10�8

10�7

10�6

10�5

10�4

10�3

10�2

10�1

100

101

d�/d

p T,j

1[p

b/G

eV]

GoSam + Sherpapp ! H + 3 jets at 13 TeV

CT14nlo, R = 0.4 anti-kT, |⌘jet| < 4.4, pT,jet = 100 GeV

LO H+3 (⇥10)

LO H+3 (veto)

LO H+3 mt,b (⇥10)

LO H+3 mt,b (veto)

LO H+3 mt (⇥10)

LO H+3 mt (veto)

preliminary preliminary

Gionata Luisoni, DIS 2017

Higgs pT


high-energy resummation of Higgs pT in gg ! H with off-shell gluons

Higgs + jet

Neumann, Williams ‘17

NLO for Higgs signalshiggs/talks/heinrich.pdf · Gudrun Heinrich Max Planck Institute for Physics, Munich NLO for Higgs signals

Documents