MadGraph5_aMC@NLO

Different cross-sections

Asked by Maksym Ovchynnikov

Hi!

I tried to generate the production cross-section of lepton-neutrino pair (including charge conjugated channel) at sqrt[s] =13 TeV and no cuts using two different processes:

generate p p > w+, (w+ > ve e+)
add process p p > w-, (w- > ve~ e-)

generate p p > ve e+
add process p p > ve~ e-

I expected that the first process is included in the second process, while the second process includes also sub-processes like
p p > w j, w > ve e+, ... Therefore the cross-section for the first process must be larger. However, it turned out that the cross-section for the first process is 1.57*10^4 pb, while for the second one it is 0.97*10^4 pb, i.e. smaller in 1.5 times.

What can be a reason for this?

Question information

Language:
English Edit question
Status:
Solved
For:
MadGraph5_aMC@NLO Edit question
Assignee:
No assignee Edit question
Solved by:
Olivier Mattelaer
Solved:
Last query:
Last reply:
Revision history for this message
Olivier Mattelaer (olivier-mattelaer) said : 		#1

Olivier Mattelaer suggests this article as an answer to your question:
FAQ #2778: “Different syntax for the same diagram leads to different cross-sections: Why?”.

Revision history for this message
Maksym Ovchynnikov (name-xxx) said : 		#2

Dear Olivier,

thank you for this clarification. However, I still not understand why the second cross-section is smaller than the first one. It is because it is unphysical (which is not expected since the width of W is very narrow), or because the second cross-section does not include the kinematic range of the invariant mass near the Ws mass (the domain is defined by the value of bwcutoff)?

Revision history for this message
Best Olivier Mattelaer (olivier-mattelaer) said : 		#3

This is just because the second syntax has more cut on the lepton (it requires ptl>10) which is not the case with the first syntax (but if you set cut_decays to True)

Cheers,

Olivier

Revision history for this message
Maksym Ovchynnikov (name-xxx) said : 		#4

Thanks Olivier Mattelaer, that solved my question.