Challenges for EFTs at Run2 LHC Veronica Sanz (Sussex) HEFT2015, - PowerPoint PPT Presentation

Challenges for EFTs at Run2 LHC Veronica Sanz (Sussex) HEFT2015, Chicago

Outline EFTs: how do we look for New Physics with them � Challenges for EFTs: precision and breakdown � Precision: NLO QCD � Breakdown: Benchmarks in extended Higgs sectors

Effective Field Theory heavy New Physics

New Physics could be heavy � Buchmuller and Wyler. NPB (86) as compared with the typical energy of the channel we look at EFT: expansion in higher- dimensional operators (HDOs) model independent � � Systematic studies � Advantages of EFT � One operator, corrs. � � Translation to thy

Expansion in inverse powers of NP scale dim6, dim8, … coupling HWW � at dim-6 Contino et al. 1303.3873 here, assuming Higgs doublet � non-linear: see talks by Merlo, Krause, Panico

How do we look for HDOs? Rates and differential distributions

New Physics induces new coupling structures of SM particles, incl the Higgs Higgs anomalous couplings Higgs anomalous couplings − 1 − 1 4 h g (1) 4 h g (1) HDOs generate HDOs generate hV V V µ ν V µ ν hV V V µ ν V µ ν HVV interactions HVV interactions − h g (2) − h g (2) hV V V ν ∂ µ V µ ν hV V V ν ∂ µ V µ ν with more with more − 1 − 1 g hV V V µ ν ˜ g hV V V µ ν ˜ V µ ν V µ ν derivatives derivatives 4 h ˜ 4 h ˜ ex. Feynman rule if mh>2mV ex. Feynman rule if mh>2mV ✓ ˆ ✓ ˆ ✓ ✓ ◆ ◆ ◆ ◆ V ( p 2 ) V ( p 2 ) s s g (1) g (1) + 2 g (2) + 2 g (2) 2 − m 2 2 − m 2 hV V m 2 hV V m 2 i η µ ν i η µ ν V V V V hV V hV V h ( p 1 ) h ( p 1 ) − ig (1) − ig (1) hV V p µ hV V p µ 3 p ν 3 p ν 2 2 V ( p 3 ) V ( p 3 ) g hV V ✏ µ ναβ p 2 , α p 3 , β g hV V ✏ µ ναβ p 2 , α p 3 , β − i ˜ − i ˜

New Physics induces new coupling structures of SM particles, incl the Higgs Higgs anomalous couplings Higgs anomalous couplings ✓ ˆ ✓ ˆ ✓ ✓ ◆ ◆ ◆ ◆ V ( p 2 ) V ( p 2 ) s s g (1) g (1) + 2 g (2) + 2 g (2) 2 − m 2 2 − m 2 hV V m 2 hV V m 2 i η µ ν i η µ ν V V V V hV V hV V h ( p 1 ) h ( p 1 ) − ig (1) − ig (1) hV V p µ hV V p µ 3 p ν 3 p ν 2 2 V ( p 3 ) V ( p 3 ) g hV V ✏ µ ναβ p 2 , α p 3 , β g hV V ✏ µ ναβ p 2 , α p 3 , β − i ˜ − i ˜ Changes in � total rates and differential information

Anomalous couplings vs EFT coefficients Gorbahn, No, VS. 1502.07352

Differential information: � channels which probe a large kinematic regime � e.g. VH and H+j q V V ∗ ATLAS-CONF-2013-079 LHC8 q 0 h

LHC8 ATLAS VH 12 simulation 10 8 c W = 0 . 1 ¯ N ev 6 ATLAS-CONF-2013-079 c W = 0 . 05 ¯ LHC8 4 2 SM 0 0 50 100 150 200 250 p T H GeV L � Ellis, VS and You. 1404.3667, 1410.7703 Feynrules -> MG5-> pythia->Delphes3 � verified for SM/BGs => expectation for EFT c W Global fit inclusive cross section is less sensitive than distribution

How bad is it? Run1 constraints one-by-one global � Ellis, VS and You. 1410.7703 stronger in classes of models � global e.g. extended Higgs sectors � Gorbahn, No, VS. 1502.07352

Best sensitivity to new physics � exploiting differential information Challenges for EFTs at Run2

LHC8 ATLAS VH 12 10 most sensitive bin: � 8 ¯ c W = 0 . 1 N ev overflow (last) bin 6 c W = 0 . 05 ¯ 4 2 SM 0 0 50 100 150 200 250 p T H GeV L At high-pT � sensitive to dynamics of new physics � breakdown of EFT To what extent can we use this bin? how far does it extend?

Generally speaking Challenges of looking at tails of distributions Precise determination � 1 Higher-order SM and EFT under control Range of validity � 2 Need of benchmarks

Precision, precision

Differential distributions Better theory calculations, � but also inclusion in a MC generator depend on cuts � need radiation and detector effects Simulation tools theory Collider � f i X L eff = Λ 2 O i simulation i observables Limit coefficients � = new physics

example: NLO QCD in VH LO vs NLO, showering effects Higgs-Z invariant mass ( pp → H Z → b ¯ b ` + ` − ) 10 − 1 SM dM VH [ fb / 25 GeV ] 10 − 2 POWHEG+PYTHIA8 POWHEG+PYTHIA8 d � MCFM NLO MCFM NLO MCFM LO MCFM LO 10 − 3 60 30 � (%) 0 − 30 − 60 150 200 250 300 350 400 450 M VH [ GeV ] � Mimasu, VS, Williams. in prep

example: NLO QCD in VH NLO QCD POWHEG+PYTHIA8 Higgs-Z invariant mass ( pp → H Z → b ¯ b ` + ` − ) 10 − 1 dM VH [ fb / 25 GeV ] 10 − 2 SM ( q ¯ SM ( q ¯ q + gg ) q + gg ) d � c W = − 0.02 c W = − 0.02 c HW = 0.015 c HW = 0.015 10 − 3 100 � BSM (%) 50 0 − 50 − 100 150 200 250 300 350 400 450 M VH [ GeV ] � Mimasu, VS, Williams. in prep alternative tool in aMC@NLO deGrande, Fuks, Mawatari, Mimasu, VS. in prep

Matching UV completions to the EFT � Gorbahn, No, VS. 1502.07352 recent paper by Brehmer, Freitas, Lopez-Val , Phlehn. 1510.03443

LHC8 ATLAS VH Where/how does the EFT break 12 10 down? depends on UV completion 8 c W = 0 . 1 ¯ N ev 6 c W = 0 . 05 ¯ 4 Need benchmarks to test the validity 2 0 of the approach 0 50 100 150 200 250 p T H GeV L Breakdown depends on loop-induced or tree-level Benchmarks: Extended Higgs sectors Gorbahn, No, VS. 1502.07352 1. Tree-level mixing: Higgs+Singlet � 2. Loop-induced EFT: 2HDMs � 3. Tree-level exchange: Radion/Dilaton

LHC8 ATLAS VH Where/how does the EFT break 12 10 down? depends on UV completion 8 c W = 0 . 1 ¯ N ev 6 c W = 0 . 05 ¯ 4 Need benchmarks to test the validity 2 0 of the approach 0 50 100 150 200 250 p T H GeV L Breakdown depends on loop-induced or tree-level

In a nutshell, we did the matching � EFT to UV models and combined EWPTs, Direct searches and Higgs limits in this framework 50 pages of gory details…

For example, for 2HDM

For example, for 2HDM Matching to EFT: unbroken phase checked the results by matching in the broken theory EWPTs limits

For example, for 2HDM Matching to EFT: unbroken phase Sensitivity � sizeable quartic couplings � or light particles Next step � quantify the EFT breakdown within these benchmarks � Mimasu, No, VS. in prep. See also, Brehmer, Freitas, Lopez-Val , Phlehn.1510.03443

Conclusion Best sensitivity to NP in EFTs requires handling differential distributions � Challenges: Precision and breakdown � Precision: push understanding of SM and EFTs at higher orders, implementation in tools for simulations � Breakdown: model-dependent question. Propose benchmarks, matching between EFT and UV models, include them in tools (e.g. loop-induced requires form- factors), quantify differences

In the Higgs basis

1000 LHC8 900 c W = − 0 . 025 ¯ 800 validity 700 (GeV) 600 VH m 500 400 300 Associated production VH 200 200 250 300 350 400 450 500 V p (GeV) T distribution m W √ c = g NP Λ NP Λ NP ' g NP ( 0.5 TeV ) Ellis, VS, You. 1404.3667

In terms of Higgs’ anomalous couplings black global fit � green one-by-one fit