SM Higgs to Z gamma decay for top contribution only

Hi,

Your syntax is ok and I tested a second equivalent syntax and got the same result.

Cheers,

Olivier

Hi Olivier

Thanks a lot for your answer. I also checked the h \to \gamma \gamma decay with top contribution only and the top contribution is much smaller than the W contribution.

Could you please help me understand the decay width of the h \to Z \gamma process considering only the top quark contribution in Madgraph? Why does it give a larger decay width for the top contribution compared to the W-boson contribution? Is there any interference going on while doing the numerical calculation?

Thanks again for your time and help.

Best
Hridoy

Hi,

I have taken some look,
First, I have found some non cancelling pole from collier method. But it seems that the issue is that madgraph ask him to not compute them (since we know they are zero), I have checked with cuttools and ninja and the three of them aggree on the result.

Now on your statement that the top does not dominates, I'm not fully sure that this is True.
If I look at this paper:
http://arxiv.org/pdf/1505.00561
they say "For the dominant top quark contribution, " (end of section 2).

But indeed from http://arxiv.org/pdf/1505.00561
I see the statement:
We note,
moreover, that in the SM the W -boson loop contribution
dominates by far over the top-loop contribution, while
those of the b-quark and other fermions are negligible

So I guess that I miss-interpret wrongly at least one paper, or that they refer to something else...

Cheers,

Olivier

Hi Olivier

I can't thank you enough for taking the time to look at this in different ways and answer my question.

Just to answer your question about the top contribution, the paper you shared http://arxiv.org/pdf/1505.00561 also says the same. If you look at Table 1 (page 13) of this paper, the authors computed all the contributions independently, and the top contribution is much smaller. In that paper, in Table 1, they showed that the W contribution is 7.86 KeV which exactly matches what Madgraph gives but the top contribution doesn't.

 Besides this, I did this calculation by hand and also double-checked the result with Package X. Every time I got that the top contribution was much smaller than what the Madgraph gives.

The reason I am worried about this decay as I am doing this same process with Z' \gamma. As It gives a much larger decay width for SM Z \gamma, it also gives a much larger decay width for Z' \gamma.

Thanks again. Please let me know if you have any suggestions.

Best
Hridoy

I have look at the code and do not spot any obvious mistake, I have also remove the R2 and UV counter-term and they are not the source of such large number. (i.e this is coming from the loop itself).

Now I have notice the following, If I check the full process (or the one without the top) everything is fine:

check h > z a / t [virt=QCD QED]
check: Samurai not available on your system; it will be skipped.
INFO: Checking crossings of process h > z a [ virt = QCD QED ] / t
INFO: Testing permutation: [1, 3, 2]
INFO: Skipping Process: h > z a [ sqrvirt = QCD QED ] / t, identical matrix element already tested
DEBUG: Cleaning temporary TMP_CHECK* check runs.
INFO: Checking lorentz transformations for process h > z a [ virt = QCD QED ] / t
DEBUG: Reusing generated output /Users/omattelaer/Documents/git_workspace/LTS/TMP_CHECK
DEBUG: Reusing generated output /Users/omattelaer/Documents/git_workspace/LTS/TMP_CHECK
DEBUG: Reusing generated output /Users/omattelaer/Documents/git_workspace/LTS/TMP_CHECK
DEBUG: Reusing generated output /Users/omattelaer/Documents/git_workspace/LTS/TMP_CHECK
DEBUG: Reusing generated output /Users/omattelaer/Documents/git_workspace/LTS/TMP_CHECK
DEBUG: Cleaning temporary TMP_CHECK* check runs.
INFO: Checking ward identities for process: h > z a [ virt = QCD QED ] / t
DEBUG: Reusing generated output /Users/omattelaer/Documents/git_workspace/LTS/TMP_CHECK
DEBUG: Cleaning temporary TMP_CHECK* check runs.
check: 3 check performed in 39 seconds
DEBUG:
Lorentz invariance results:
h > z a
Transformation name Value Relative diff. Result
Original evaluation 1.0548553536e-01
Z-axis boost 1.0548553536e-01 4.7361978053e-15 Passed
X-axis boost 1.0548553536e-01 6.8280185026e-14 Passed
Y-axis boost 1.0548553536e-01 2.3680989026e-14 Passed
Z-axis pi/2 rotation 1.0548553536e-01 0.0000000000e+00 Passed
Z-axis pi/4 rotation 1.0548553536e-01 1.9734157522e-15 Passed
Summary: passed
Gauge results:
Process [virt=QCD QED] matrix BRS ratio Result
h > z a 1.0548553536e-01 2.4267115288e-30 2.3005159147e-29 Passed
Summary: 1/1 passed, 0/1 failed
Process permutation results:
Process [virt=QCD QED] Min element Max element Relative diff. Result
h > z a * No permutations, process not checked *
Summary: 0/0 passed, 0/0 failed
Not checked processes: Process: h > z a [ virt = QCD QED ] / t

But if I check only the top loop, then this break gauge and lorentz invariance:

MG5_aMC>check h > t > z a [virt=QCD QED]
check: Samurai not available on your system; it will be skipped.
INFO: Checking crossings of process h > t > z a [ virt = QCD QED ]
INFO: Testing permutation: [1, 3, 2]
INFO: Skipping Process: h > t > z a [ sqrvirt = QCD QED ], identical matrix element already tested
DEBUG: Cleaning temporary TMP_CHECK* check runs.
INFO: Checking lorentz transformations for process h > t > z a [ virt = QCD QED ]
DEBUG: Reusing generated output /Users/omattelaer/Documents/git_workspace/LTS/TMP_CHECK
DEBUG: Reusing generated output /Users/omattelaer/Documents/git_workspace/LTS/TMP_CHECK
DEBUG: Reusing generated output /Users/omattelaer/Documents/git_workspace/LTS/TMP_CHECK
DEBUG: Reusing generated output /Users/omattelaer/Documents/git_workspace/LTS/TMP_CHECK
DEBUG: Reusing generated output /Users/omattelaer/Documents/git_workspace/LTS/TMP_CHECK
DEBUG: Cleaning temporary TMP_CHECK* check runs.
INFO: Checking ward identities for process: h > t > z a [ virt = QCD QED ]
DEBUG: Reusing generated output /Users/omattelaer/Documents/git_workspace/LTS/TMP_CHECK
DEBUG: Cleaning temporary TMP_CHECK* check runs.
check: 3 check performed in 34 seconds
DEBUG:
Lorentz invariance results:
h > z a
Transformation name Value Relative diff. Result
Original evaluation 4.6419499646e-02
Z-axis boost 5.3881725133e-02 1.4879629842e-01 Failed
X-axis boost 5.7158945911e-02 2.0736836138e-01 Failed
Y-axis boost 4.8708884340e-02 4.8132525704e-02 Failed
Z-axis pi/2 rotation 4.6419499646e-02 1.4948230715e-16 Passed
Z-axis pi/4 rotation 4.6419499646e-02 4.4844692145e-16 Passed
Summary: failed
Gauge results:
Process [virt=QCD QED] matrix BRS ratio Result
h > z a 4.6419499646e-02 6.2510771709e+01 1.3466489769e+03 Failed
Summary: 0/1 passed, 1/1 failed
Failed processes: h > z a
Process permutation results:
Process [virt=QCD QED] Min element Max element Relative diff. Result
h > z a * No permutations, process not checked *
Summary: 0/0 passed, 0/0 failed
Not checked processes: Process: h > t > z a [ virt = QCD QED ]

So clearly they are issue with the top only diagram which likely means that some counter-term (or ghost?) are not correctly handle in that case.
I would suggest to run the same check for your model to see if the top contribution alone makes sense or not.

Cheers,

Olivier

HI Olivier

Thanks a lot. I just followed your command and it looks like the Lorentz invariance is broken but passed the gauge invariance.

Here I have attached the output I got.

check h > t > zb a [virt=QCD]
check: Samurai not available on your system; it will be skipped.
INFO: Checking process h > t > zb a [ virt = QCD ] in unitary gauge
INFO: Checking process h > t > zb a [ virt = QCD ] in feynman gauge
INFO: Checking crossings of process h > t > zb a [ virt = QCD ]
INFO: Testing permutation: [1, 3, 2]
INFO: Skipping Process: h > t > zb a [ sqrvirt = QCD ], identical matrix element already tested
INFO: Checking lorentz transformations for process h > t > zb a [ virt = QCD ]
INFO:Checking ward identities for process: h > t > zb a [ virt = QCD ]
Lorentz invariance results:
h > zb a
Transformation name Value Relative diff. Result
Original evaluation 3.4679339926e-02
Z-axis boost 1.1858333257e-01 1.0949044706e+00 Failed
X-axis boost 1.1449755092e-01 1.0701149560e+00 Failed
Y-axis boost 8.1595934981e-02 8.0699177176e-01 Failed
Z-axis pi/2 rotation 3.4679339926e-02 0.0000000000e+00 Passed
Z-axis pi/4 rotation 3.4679339926e-02 0.0000000000e+00 Passed
Summary: failed
Gauge results:
Process [virt=QCD] matrix BRS ratio Result
h > zb a 3.4679339926e-02 3.0452333218e+03 8.7811167350e+04 Failed
Summary: 0/1 passed, 1/1 failed
Failed processes: h > zb a
Gauge results (switching between Unitary/Feynman/axial gauge):
Process [virt=QCD] Unitary Feynman Relative diff. Result
h > zb a 3.4679339925e-02 3.4679339926e-02 2.9630921477e-11 Passed
Summary: 1/1 passed, 0/1 failed
Process permutation results:
Process [virt=QCD] Min element Max element Relative diff. Result
h > zb a * No permutations, process not checked *
Summary: 0/0 passed, 0/0 failed
Not checked processes: Process: h > t > zb a [ virt = QCD ]

Thanks again for your help.

Best
Hridoy

Hi Olivier

Please let me know if you have any ideas/suggestions about how to solve this issue.

Thanks again.

Best
Hridoy