MadGraph5_aMC@NLO

ttbarZ cross section conflict mc12 and mc11 and madgraph

Asked by Arash

I have noticed that mc12_8TeV. 119355.MadGraphPythia_AUET2BCTEQ6L1_ttbarZ.SUSY_NTUP.e1352_s1499_s1504_r3658_r3549_p1032
reports the cross section as about 67.69 fb but when I use http://madgraph.hep.uiuc.edu/ to find the ttbarZ cross section I find 144 fb.
Could you tell me why there is a difference and which one is the correct one?

Thanks

Question information

Language:
English Edit question
Status:
Solved
For:
MadGraph5_aMC@NLO Edit question
Assignee:
No assignee Edit question
Solved by:
Arash
Solved:
Last query:
Last reply:
Revision history for this message
Johan Alwall (johan-alwall) said : 		#1

Hello Arash,

Can you please post the banner (.lhe file header) for the mc12_8TeV. 119355.MadGraphPythia_AUET2BCTEQ6L1_ttbarZ.SUSY_NTUP.e1352_s1499_s1504_r3658_r3549_p1032 run?

Thanks,
Johan

Revision history for this message
Arash (akhazr) said : 		#2

Hello Johan,

In athena_stdout.txt the file madgraph.119355.MadGraph_ttbarZ._00001.events.tar.gz
is linked to events.lhe
it is at folk.uio.no/borgeg/tmp10/

Revision history for this message
Johan Alwall (johan-alwall) said : 		#3

Hello Arash,

Thanks for the file. In that file, the cross section is reported as 146 fb:
<MGGenerationInfo>
# Number of Events : 20000
# Integrated weight (pb) : .14631E+00
# Truncated wgt (pb) : .00000E+00
# Unit wgt : .73155E-05
</MGGenerationInfo>
(and also in the <init> block).

 Where do you find the number 67.69 fb?

All the best,
Johan

Revision history for this message
Arash (akhazr) said : 		#4

Hell Johan,

I found 67.69 reported at http://ami.in2p3.fr/opencms/opencms/AMI/www/
totalEvents
400000
dataType

PDF

CTEQ6L1 - LO with LO alpha_s
TransformationPackage
17.2.3.6
ecmEnergy
8000
prodsysStatus

ALL EVENTS AVAILABLE
AtlasRelease

I suppose there has been a mistake there?

Thanks

Regards

Arash

Revision history for this message
Arash (akhazr) said : 		#5

Hi,

I found the answer from susy d3pd group:

The ttbarZ process was produced in two separate samples, one with zero additional partons (ttbarZ 119355) and another with one additional parton (ttbarZj 119356). These samples were produced with MLM matching applied and must *both* be combined for analysis.

The MLM matching procedure applies a filter to the (exclusive) ttbarZ sample in order to avoid overlaps with the (inclusive) ttbarZj sample. Hence there is a filtering efficiency of ~50% which explains the discrepancy you observe.

If the ttbarZ and ttbarZj samples are combined the total cross section is 67.7+87.3=155 fb which is closer to the value you get from madgraph of 144 fb. Also note that the NLO estimate of the ttbarZ cross section is 208 fb hence a k-factor of 1.3 should also be applied to these samples.

Thanks

Regards

Arash

Revision history for this message
Arash (akhazr) said : 		#6

Hi,

I found the answer from susy d3pd group:

The ttbarZ process was produced in two separate samples, one with zero additional partons (ttbarZ 119355) and another with one additional parton (ttbarZj 119356). These samples were produced with MLM matching applied and must *both* be combined for analysis.

The MLM matching procedure applies a filter to the (exclusive) ttbarZ sample in order to avoid overlaps with the (inclusive) ttbarZj sample. Hence there is a filtering efficiency of ~50% which explains the discrepancy you observe.

If the ttbarZ and ttbarZj samples are combined the total cross section is 67.7+87.3=155 fb which is closer to the value you get from madgraph of 144 fb. Also note that the NLO estimate of the ttbarZ cross section is 208 fb hence a k-factor of 1.3 should also be applied to these samples.

Thanks

Regards

Arash