This meeting is dedicated to physics modeling issues in the Wmass measurement at the LHC. It is part of the 2017 TH Institute on
"LHC and the Standard Model: Physics and Tools",
Minutes and action items of the MW topical meeting on June 22, 2017 at
- write up a short summary of methods used in the ATLAS analysis (Maarten)
Most of the discussion was about understanding the master formula for
the physics modeling on slides 7,8 of Stefano's talk and how theory
uncertainties propagate to the W masss sytematics.
It was also asked what would happen if one theory prediction is used for
all components instead of this 'composite' model?
- initiate a quantitative study of the correlation/decorrelation of QCD
uncertainties in W and Z production (pT distribution in particular)
Are there theory arguments for a certain procedure, for instance, it was
suggested that there are good arguments for keeping the scale variation
for light quarks and gluons correlated in the W/Z ratio but not for
Another suggestion was to revisit the definition of the ratio, e.g., to
study the effect of kinematic differences to first 'align the Sudakovs'
and then take the ratio.
It was emphasized that NNLO+NNLL uncertainties are still relatively
large in case of small qT even in normalized distributions,
that there is a 5% scale uncertainty in the ratio (fully correlated),
and that interestingly the observed behaviour of the ratio
agrees better with NNLL when the parton evolution is done at LO
(see Massimiliano's talk)
- further studies on understanding the origin of the hardening of the pTW
distribution (wrt pTZ) when going to NNLL?
Shall we plan a dedicated comparison of DYRes, Resbos, Cute, DIRE,
Sherpa, and Geneva?
See remark above and Massimiliano's slides showing a comparison of DYres,
Resbos, CuTe with Phythia 8 AZ
- role of heavy quarks
discussion about the role played by collinear PDFs and sensitivity
to improvements in the description of the heavy quark kinematics
(follow up on Alessandro's study on b-quark mass effects,
improved charm treatment? )
After a discussion with Emanuele Re, it seems that the Les Houches study
will mostly focus on the different approaches implemented in Parton
Shower tools to handle the transition from bottom to g-> b bbar
in the backward evolution
and the impact of different choices on the ptZ prediction
Frank Krauss suggested to check the y dependence of flavour decomposition,
(already in progress) and the effect of flavour-blind kT kick to partons.
Frank Krauss had a number of ideas for further studies of parton shower
uncertainties (e.g. is the scale in alphas in Pythia varied) and noted that
Sherpa has only one parameter for tuning. There are now two Sherpa PS (Dire)
and it was suggested to compare them.
-Z+jet at NNLO matched to analytic resummation is work in progress (see Aude
Gehrmann's talk at the TH Institute, with P.F. Monni). We also need
W+jet at NNLO(+resummation) (timescale?).
-Maria's talk on Z pT and PDFs mainly concentrated on the impact of the
inclusion of large pT Z data. It was pointed out that to be able to
include the experimentally very precise small pt Z data one needs to
include resummation in the predictions. Other issues are how to estimate
the size of NP corrections and how to treat the relatively large
statistical uncertainty of the Monte Carlos providing the higher-order
predictions. Also, it was asked if we need to worry about power
corrections at this level of precision.