Z boson fusion

Asked by Dimitris Athanasakos on 2021-03-08

Hi all,

I am trying to reproduce the results of this paper (https://arxiv.org/abs/2008.12204) and I am having trouble with the Z boson fusion channels (all of them actually).

For example, consider the process mu- mu+ > mu- mu+ h. There are only two diagrams that contribute: one s-channel process and the zbf channel. When launching the process, the cross-section seem to scale as 1/s (as if it was a pure s-channel process, which is not). To make matters worse, this problem persists even if I forcefully ignore the s-channel contribution (using the $$z command).
In particular, I am getting for s=3 Tev: σ=0.6 fb and for s=30 Tev: σ= 10^(-4) fb, while the authors are getting completely different results (see table 1 of the paper).
In general I am having trouble recreating any cross-section that includes a Z boson fusion channel (for example mu- mu+ > mu- mu+ h h) and I can't figure why. ( for the wbf channels everything works as expected)

Note: I am only using Madgraph v2.9.2 (I ran the process with both python2 and python3, and obviously it didn't change anything)

Question information

Language:
English Edit question
Status:
Solved
For:
MadGraph5_aMC@NLO Edit question
Assignee:
No assignee Edit question
Solved by:
Olivier Mattelaer
Solved:
2021-03-10
Last query:
2021-03-10
Last reply:
2021-03-08

Hi,

But your process is not in Table 1 so I do not understand

Table 1 is currently not possible to generate (within the EVA approximation) within MG5aMC.
This is currently work in progress (we are actually validating our implementation against the result of that paper)

> Note: I am only using Madgraph v2.9.2 (I ran the process with both python2 and python3, and obviously it didn't change anything)

For such type of process 2.9.x is recomended since previous version do not converge correctly (see https://arxiv.org/pdf/2102.00773.pdf <https://arxiv.org/pdf/2102.00773.pdf> in particular Table 7 page 12, there we only compare the speed difference
(estimated to 35 thousands time faster) but the estimated cross-section is actually wrong within 2.8.x version of the code as noted in the text)

Cheers,

Olivier

> On 8 Mar 2021, at 19:55, Dimitris Athanasakos <email address hidden> wrote:
>
> New question #695924 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/695924
>
> Hi all,
>
> I am trying to reproduce the results of this paper (https://arxiv.org/abs/2008.12204) and I am having trouble with the Z boson fusion channels (all of them actually).
>
> For example, consider the process mu- mu+ > mu- mu+ h. There are only two diagrams that contribute: one s-channel process and the zbf channel. When launching the process, the cross-section seem to scale as 1/s (as if it was a pure s-channel process, which is not). To make matters worse, this problem persists even if I forcefully ignore the s-channel contribution (using the $$z command).
> In particular, I am getting for s=3 Tev: σ=0.6 fb and for s=30 Tev: σ= 10^(-4) fb, while the authors are getting completely different results (see table 1 of the paper).
> In general I am having trouble recreating any cross-section that includes a Z boson fusion channel (for example mu- mu+ > mu- mu+ h h) and I can't figure why. ( for the wbf channels everything works as expected)
>
> Note: I am only using Madgraph v2.9.2 (I ran the process with both python2 and python3, and obviously it didn't change anything)
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Hi Olivier,

Now I am the one that's confused.

Below equation (2.1) the authors imply that in table 1 what's depicted is the cross-sections for the Higgs production from a mu- mu+ collider and there are two ways to achieve that through VBF, either WW->h (so the actual process is mu- mu+ > vm vm~ h) or ZZ->h (mu- mu+> mu- mu+ h ). Also when running the process: mu- mu+ > vm vm~ h, I reproduce the results in the row: WW->h (with an accuracy of 5%). Same for WW->H H.

Hi,

I know but you should also have photon Z, Z photon and photon-photon (and their interference).
Additionally in both computation you should have a scale choice. In the paper, you have the scale of the PDF which can be interpret as a maximum pT cut on the lepton.
While with the syntax that you use, you likely need a minimal pt cut.

This is actually the point raised by the paper /2008.12204 <https://arxiv.org/abs/2008.12204> compare to our paper 2005.10289 <https://arxiv.org/abs/2005.10289> where we use the same syntax as you, that we are actually cutting away the colinear part of the phase-space.

Cheers,

Olivier

> On 8 Mar 2021, at 21:45, Dimitris Athanasakos <email address hidden> wrote:
>
> Question #695924 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/695924
>
> Dimitris Athanasakos posted a new comment:
> Hi Olivier,
>
> Now I am the one that's confused.
>
> Below equation (2.1) the authors imply that in table 1 what's depicted
> is the cross-sections for the Higgs production from a mu- mu+ collider
> and there are two ways to achieve that through VBF, either WW->h (so the
> actual process is mu- mu+ > vm vm~ h) or ZZ->h (mu- mu+> mu- mu+ h ).
> Also when running the process: mu- mu+ > vm vm~ h, I reproduce the
> results in the row: WW->h (with an accuracy of 5%). Same for WW->H H.
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Thanks Olivier Mattelaer, that solved my question.