Choose timezone
Your profile timezone:
VH Cross Section Meeting
→
Europe/Zurich
CERN
CERN
, , ,
VH production cross section meeting
===================================
Indico page
https://indico.cern.ch/conferenceDisplay.py?confId=242565
Theoretical developments
------------------------
* Update for the predictions of total cross sections
Stefan reports on plans for an update of predictions that will be documented in Yellow Report 3 (YR3):
Robert Harlander et al. are running vh@nnlo to produce state-of-the-art numbers based on
- NNLO QCD corrections
-> The non-Drell-Yan-like parts (%-level) were not available for YR1
- gg->ZZ at LO and NLO
-> The NLO part (3-5%) was not available for YR1
- NLO EW corrections
* Update for the predictions of differential cross sections
Stefan again reports on plans for updated predictions for YR3:
Giancarlo + the HAWK team are comining the following contributions
- NNLO QCD predictions based on Drell-Yan-like corrections (Giancarlo)
-> NNLO part new for ZH, not yet available YR2
- relative EW corrections from HAWK (Alexander)
Experimental side progress
--------------------------
Present situation in ATLAS
https://indico.cern.ch/getFile.py/access?contribId=1&resId=0&materialId=slides&confId=242565
* Electroweak corrections
Jason Nielsen with the input from Alexander Mueck and the HAWK team derived now continous EW corrections as a function of pT(W). These corrections are sizeble and applied on top of the ATLAS analysis. We discussed whether the use of PHOTOS in the parton shower used for the signal Monte Carlo used in the exp analysis could double count part of this correction. Stefan clarified this is not the case, as most of the correction comes from non-factorizable contributions, while the residual dependence on the way real photon emission is treated (different schemes for the computation) is small.
* QCD corrections
Here more work is needed from the experimental side, to profit from the precision of the parton-level computation.
Infact, comparisons between different LO (Pythia6 with two different LO PDFs/UE tunes, Pythia8)and NLO (Herwig++ with Powheg method) parton shower MC generators show significant differences (at the level of 5-10% depending on pT(W/Z)), as shown in Michiel Sanders talk:
https://indico.cern.ch/getFile.py/access?contribId=0&resId=0&materialId=slides&confId=242565
(this is for WH, ZH will follow hopefully soon)
Ideally, to exploit the precision of the NNLO differential computation we need to disentangle ME and PDF effects from shower and hadronization model, and UE tunes effects.
Giancarlo commented he has a setup where he compares the NNLO parton level computation with MC@NLO, and he finds good agreement. This is not part of the next YR3, but is going to be published soon. He strongly suggests to only use NLO based PS MC for the comparisons.
In order to try to understand what is impacting our selection efficiency, possible next steps could be:
* ATLAS adds MC@NLO (or aMC@NLO) to his comparison [do the two include the NLO Higgs decay?]
* Giancarlo Ferrera and collaborators will also add Herwig++ to his comparison (this includes the NLO Higgs decay)
* the two setup could then be compared and hopefully additional effects from shower/hadronization model, UE tunes understood and disentangled from the rest.
It would be also useful for the experiments to have the ntuples of the NLO computation (NLO to start with, NNLO maybe later since much more time consuming):
* nominal NLO production+ NLO decay
* same with ren. scale and fact scale up and down
* if possible, same also with PDF variations
Giancarlo Ferrera has already provided very useful ntuples in the past to ATLAS (David Lopez working on this), but to my knowledge a complete and consistent set is still missing.
Work on W+b background process
Andrea Rizzi suggested to start working also within this WG on comparing different MC generators for the W+bb process. In particular, he mentions data vs MC discrepancies for W+1 b-jet at high pT. Will follow up if we find someone to work on this (afaik ATLAS also started internally some comparisons). In this case we should probably ask some theorists working on W+b as well.
Next meetings
We decided to meet again in around a month, starting a doodle poll for the next meeting in 3 weeks, after some progress has being made.
===================================
Indico page
https://indico.cern.ch/conferenceDisplay.py?confId=242565
Theoretical developments
------------------------
* Update for the predictions of total cross sections
Stefan reports on plans for an update of predictions that will be documented in Yellow Report 3 (YR3):
Robert Harlander et al. are running vh@nnlo to produce state-of-the-art numbers based on
- NNLO QCD corrections
-> The non-Drell-Yan-like parts (%-level) were not available for YR1
- gg->ZZ at LO and NLO
-> The NLO part (3-5%) was not available for YR1
- NLO EW corrections
* Update for the predictions of differential cross sections
Stefan again reports on plans for updated predictions for YR3:
Giancarlo + the HAWK team are comining the following contributions
- NNLO QCD predictions based on Drell-Yan-like corrections (Giancarlo)
-> NNLO part new for ZH, not yet available YR2
- relative EW corrections from HAWK (Alexander)
Experimental side progress
--------------------------
Present situation in ATLAS
https://indico.cern.ch/getFile.py/access?contribId=1&resId=0&materialId=slides&confId=242565
* Electroweak corrections
Jason Nielsen with the input from Alexander Mueck and the HAWK team derived now continous EW corrections as a function of pT(W). These corrections are sizeble and applied on top of the ATLAS analysis. We discussed whether the use of PHOTOS in the parton shower used for the signal Monte Carlo used in the exp analysis could double count part of this correction. Stefan clarified this is not the case, as most of the correction comes from non-factorizable contributions, while the residual dependence on the way real photon emission is treated (different schemes for the computation) is small.
* QCD corrections
Here more work is needed from the experimental side, to profit from the precision of the parton-level computation.
Infact, comparisons between different LO (Pythia6 with two different LO PDFs/UE tunes, Pythia8)and NLO (Herwig++ with Powheg method) parton shower MC generators show significant differences (at the level of 5-10% depending on pT(W/Z)), as shown in Michiel Sanders talk:
https://indico.cern.ch/getFile.py/access?contribId=0&resId=0&materialId=slides&confId=242565
(this is for WH, ZH will follow hopefully soon)
Ideally, to exploit the precision of the NNLO differential computation we need to disentangle ME and PDF effects from shower and hadronization model, and UE tunes effects.
Giancarlo commented he has a setup where he compares the NNLO parton level computation with MC@NLO, and he finds good agreement. This is not part of the next YR3, but is going to be published soon. He strongly suggests to only use NLO based PS MC for the comparisons.
In order to try to understand what is impacting our selection efficiency, possible next steps could be:
* ATLAS adds MC@NLO (or aMC@NLO) to his comparison [do the two include the NLO Higgs decay?]
* Giancarlo Ferrera and collaborators will also add Herwig++ to his comparison (this includes the NLO Higgs decay)
* the two setup could then be compared and hopefully additional effects from shower/hadronization model, UE tunes understood and disentangled from the rest.
It would be also useful for the experiments to have the ntuples of the NLO computation (NLO to start with, NNLO maybe later since much more time consuming):
* nominal NLO production+ NLO decay
* same with ren. scale and fact scale up and down
* if possible, same also with PDF variations
Giancarlo Ferrera has already provided very useful ntuples in the past to ATLAS (David Lopez working on this), but to my knowledge a complete and consistent set is still missing.
Work on W+b background process
Andrea Rizzi suggested to start working also within this WG on comparing different MC generators for the W+bb process. In particular, he mentions data vs MC discrepancies for W+1 b-jet at high pT. Will follow up if we find someone to work on this (afaik ATLAS also started internally some comparisons). In this case we should probably ask some theorists working on W+b as well.
Next meetings
We decided to meet again in around a month, starting a doodle poll for the next meeting in 3 weeks, after some progress has being made.
There are minutes attached to this event.
Show them.
