Jet binning uncertainties in ggF

Europe/Zurich
Vidyo meeting

Vidyo meeting

Bruce Mellado Garcia (University of the Witwatersrand), Daniele Del Re (Universita e INFN, Roma I (IT)), Frank Tackmann, Gavin Salam (CERN)
ACTION ITEMS:
==========
- in order to compare jet bin uncertainties the three group involved need to 
   have same starting point (e.g. inclusive cross section) in order to 
   compare results
- experiments should think about parameterization shown by Jon in the first 
   part of his talk, since it is very general. To be also discussed in the combination
   group in view of the ATLAS-CMS combination (Summer).
- CMS should implement the efficiency method as done in ATLAS
- truth instead of reco for jets quantities: ATLAS - CMS discussion would be 
   very helpful, especially for the ATLAS-CMS combination
 
 
DETAILED MINUTES
=============
 
Theory talk I (Thomas Becher)
------------------------
- Latest results on resummation for the jet veto cross section
 
Questions and comments
- Gavin: important to check on assumptions when comparing different methods. 
   Each group has its own preferences. Important have feedback from experiments 
   too.
- Daniele: please repeat calculation for R = 0.5 too (CMS)
- Pietro: how does uncertainty compare to YR3? 
   * R Frank T.: reduced by 50%. In next talks
 
Theory talk II (Jonathan Walsh)
------------------------
- Include resummation extending ST method
 
Questions and comments
- Bruce: what is rho?
   * R: migration between the 2-jet bin and the inclusive 1-jet bin.
- Daniele: how do you choose threshold p_T^off?
   * R: optimized to have validity of assumptions on both regimes (low and high
     pt). Threshold varied, as in left figure of slide 14, good stability
- Thomas: the value of total cross section for us is very similar to the 
   Higgs WG, while here the value looks quite different. 
- Thomas: you stress that you are able to address all correlations but reading 
   the paper seems that you still have assumptions. 
   * R: tend to agree. Present estimates come from our best guesses.
 
Theory talk III (Pier Francesco Monni)
-----------------------------
- Jet-bin uncertainties using jet efficiencies
 
Questions and comments
- happy about consistency among different methods?
   *  R (Gavin): first, we need to agree on the total cross section. Good 
      to start from Higgs cross section with uncertainty otherwise
      difficult to get agreement.
   *  (Petriello): at some point we are going to have N^3LO results on total
     cross section, not clear what it is going to happen in a year.
   *  (Frank T.): if YR3 number has to be used, easy to rescale our result. 
- pi2 resummation: what is the uncertainty? 
   * R (Frank T.) +- 6.9%, very close to the YR (+7.2%/-7.8%)
 
 
Use of uncertainties from latest resummed Higgs predictions

in experimental analyses (Dag Gillbert) 
------------------------------------------------
- check improvement in new resummed Higgs jet bin uncertainties 
   wrt fixed order ST in ATLAS analysis
- importance of using parameterization vs truth jet quantities vs
   reco ones
- impact of resummation uncertainties on Higgs pt spectrum.
   how to deal with correlation?
 
Questions and comments
- inclusive Higgs pt effect? not clear why, you should not get any 
   uncertainty due to jets
   * R: you have an uncertainty even if you dont' cut on jet pt
   since jet pt and Higgs pt are correlated
   * R (Rei): higgs pt uncertainty, and correlation with jet pt:  
     No theorists know how to deal with it. Important to diagonalize 
     correlation matrix
   * R (Frank T.): you can compare with what you get with the standard 
     uncertainty on Higgs pt spectrum
- for theorists: are these exclusive cross section vs nbins 
   compared to theory useful?
   * R (Frank T.): very important.
- Giovanni P.: how are the plots in slide 9 done?
   * R: weight 0-jet, 1-jet, 2-jet category by the uncertainty parameters
- Daniele: when modeling vs true jet pt, do you perform a real unfolding? 
   * R: yes but without considering model dependence 
- Gavin: slide 6 shows that the total cross section issue is 
   important to get consistency between the methods.
   (a lot of discussion on how to proceed) 
- Given the arbitrariness in correlation prescriptions, and also
   because some theory predictions are likely to improve significantly
   in the coming year, there might be value in the experiments
   exploring how to present their results so that extracted mu-values
   and couplings can be easily updated to account for new theory  
   predictions (e.g. maybe along the lines of 1401.0080).
 
 
ggF jet bin uncertainties in ATLAS H→WW (David Hall)
-------------------------------------------
- impact of of new resummed Higgs jet bin uncertainties 
   with jet veto efficiency method in ATLAS analysis
- comparison with combined inclusive method
 
 
Treatment of correlations for analyses with multiple 2-jet categories (G. Petrucciani)
-------------------------------------------------------------------
- question is: how do we deal with correlations in combination when 
   considering multiple categories with two jets? (e.g. two VBF-like categories 
   or VBF and VH)
 
Questions and comments
- no problem to include correlation in VBF categories if ST approach (YR3) is used.
- Frank T.: for VH and VBF maybe better to have them correlated,
   in particular if the pt jet drives the uncertainty.
 
 
There are minutes attached to this event. Show them.
    • 15:00 15:20
      Theory I 20m
      Speaker: Thomas Becher (University of Bern)
      Slides
    • 15:30 15:50
      Theory II 20m
      Speaker: Jon Walsh (LBNL, UC Berkeley)
      Slides
    • 16:00 16:20
      Use of uncertainties from latest resummed Higgs predictions 
in experimental analyses 20m
      Speaker: Dag Gillberg (CERN)
      Slides
    • 16:30 16:50
      Theory III 20m
      Speakers: Giulia Zanderighi, Pier Francesco Monni (ITP, UZH Zuerich)
      Slides
    • 17:00 17:20
      ggF jet bin uncertainties in H->WW 20m
      Speaker: David Christopher Hall (University of Oxford (GB))
      Slides
    • 17:25 17:35
      Treatment of correlations for analyses with multiple 2-jet categories? 10m
      Speakers: Biagio Di Micco (Unknown), Dr Biagio Di Micco (Universita' degli Studi di Roma Tre e Istituto Nazionale di Fisica Nucleare (INFN)), Giovanni Petrucciani (CERN)
      Slides