RRT: Higgs self-coupling and di-Higgs

Wednesday 21 May 2025, 23:00 → 23:59 Europe/Zurich

https://desy.zoom.us/j/69434264047

Meeting ID: 694 3426 4047
Passcode: 314159

23:00 → 23:15 pre-discussion of talks of Aidan & Jenny at meeting with PPG EW

Speakers: Aidan Robson (University of Glasgow (GB)), Jenny List (Deutsches Elektronen-Synchrotron (DE))

• main point to convey is that the new 550 GeV numbers are on a solid basis and even conservative
• Jenny will circulate draft slides on weekend...

23:15 → 23:25 do we need g_HHH vs g_HHVV ?

Speaker: Caterina Vernieri (SLAC National Accelerator Laboratory (US))

HHH_HHVV_Toyama.pdf

• upshot: Junping’s talk from 2016 illustrates the qualitative effects 
• once we have a full re-analysis of ZHH, we can redo these fits or even try look at differential information 

23:25 → 23:35 can / should we update the parametric uncertainties on trilinear from ZHH in SMEFT?

Speakers: Junping Tian (University of Tokyo), Michael Peskin

• Michael gives it a try :)
• since now 10% level precision, linear approximation is sufficient, so can include trilinear from ZHH in LO SMEFT fit

23:35 → 23:59 what’s our take on trilinear self-coupling from single-Higgs, eett operators and all that?

Speakers: Junping Tian (University of Tokyo), Michael Peskin

Way forward:

• • Michael & Junping get uncertainties for 365 GeV only and for  LCs on cross-section an A_FB from Victor Mirailles to recheck our estimate of the impact of eett
  • then seek email discussion with Tevong You et al about more details of how they implemented eett, and which exp precisions they assumed
  • then see whether meeting is necessary/ beneficial
    
    

More detailed items on discussion:

• MEP on paper by Tevong You et al (https://arxiv.org/abs/2503.13719)
  • very many observables
  • constrain all flat directions
  • which data used to constrain eett operators to which level?
  • flat directions in ttbar cannot be resolved with one energy?
• JRR on SMEFit paper (https://arxiv.org/abs/2504.05974)
  • with 365 GeV gets better by factor 2 wrt HL-LHC (30% -> 15%)
• Maggie: theory uncertainties on inputs (!) neglected everywhere, but larger impact expected for loop-level interpretations 
• Junping:
  • https://arxiv.org/abs/2503.13719 quite consistent with our expectations
  • scenario with eett ops is U(2)^5, the improvement from 60% (v1) to 29% (v2) driven by factor two improvement of HL-LHC
  • the “3rd gen dominance” scenario does NOT (!) contain eett operators
  • note added in v2 claiming that the SMEFit paper doesn’t include eett ???
• Georg on SMEFit paper
  • If one wants to determine a coupling from a global fit, one has to allow its central value to vary. If one knows or assumes the correct central value from the start, I don’t think that the result of such a fit can be called the `determination’ of this parameter. More specifically for the case of kappa_lambda, even if one is only interested in the sensitivity for the case where kappa_lambda = 1 is realised in nature, I am not convinced that forcing kappa_lambda = 1 to be realised exactly and considering only the bounds on the Wilson coefficients around zero is sufficiently general. Looking into the plots on slide 6 of Alejo’s talk (Figs. 1 and 2 of their paper) you can see that the variations of the Wilson coefficients are sharp lines that are all exactly at zero for delta-mu = 0. This is not meant to be the result of their fit, but in general I would expect that the experimental and theoretical uncertainties should have the effect that the central values of the Wilson coefficients should not be exactly at zero even if one uses the pure SM as input. 
  • As far as I am aware, Jenny, myself and our collaborators are the only people who carry out SMEFT fits where the `true’ values of kappa_lambda differ from 1. 
  •  
    In the paper by Alejo et al it is stated `Theory uncertainties are neglected for the FCC-ee observables’. I guess this means all observables, including for instance the EWPOs. I think this assumption could have a significant impact. 
  •  

    They also claim in the paper that it were difficult to come up with a BSM scenario yielding a large deviation in kappa_lambda while the effects on the single Higgs observables are small. We have shown for various models of extended HIggs sectors that exactly this happens as a generic feature, and what they write also contradicts the SMEFT results from McCullough et al. 

  • Regarding the results that they are claiming, if one believes the +-30% accuracy on kappa_lambda from the HL-LHC projection, it is obvious that the accuracy should somewhat improve if one performs a global fit including all FCC-ee data. Their factor 2 improvement from 30% to 15% looks too optimistic to me. But I guess in order to make a more quantitative statement one needs to carry out such a fit taking into account all relevant effects.