Question on beam energy spread

FCCAnalysis
1

Hello,

I am wondering about the implementation of the beam energy spread. When I open for instance a ZH file (from Pythia), and I look at the z momentum of the initial electrons, I get something like this:
root [12] events->Scan(“Particle.PDG:Particle.generatorStatus:Particle.momentum.z:Particle.mass”,“Particle.generatorStatus==4”)

***********************************************************************
*    Row   * Instance * Particle. * Particle. * Particle. * Particle. *
***********************************************************************
*        0 *        0 *        11 *         4 * 119.88084 * 0.0005109 *
*        0 *        1 *       -11 *         4 * -119.8808 * 0.0005109 *
*        1 *        0 *        11 *         4 * 120.17902 * 0.0005109 *
*        1 *        1 *       -11 *         4 * -120.1790 * 0.0005109 *
*        2 *        0 *        11 *         4 * 119.92214 * 0.0005109 *
*        2 *        1 *       -11 *         4 * -119.9221 * 0.0005109 *
*        3 *        0 *        11 *         4 * 119.92800 * 0.0005109 *
*        3 *        1 *       -11 *         4 * -119.9280 * 0.0005109 *

So the energy is spread (good !), but it seems that the event is always in the center-of-mass frame, as pz(e+)+pz(e-) = 0.
So isn’t the (arguably small) boost along z in the lab frame not taken into account ?

Cheers,
Nicolas

2

Salut Nicolas,
Yes indeed, we also noted recently oddities with the BES in Pythia, among which the one you mention. For processes for which the BES matters, we’ll use other MCs, as Whizard, for which we see no issue with the BES treatment. For ZH, a Whizard sample has already been produced for Z->mumu (wzp6_ee_mumuH_ecm240 ) and others will follow.
Cheers,
E.

3

As pointed out by @nmorange to me privately, if we specify PDF:lepton=off (it is on by default) the returned cross section does make more sense with BES than it used to be when on. It might be interesting to rerun imbalance with a test sample with BES, ISR and PDF:lepton=off

4

Salut Clément,
It does not change, no. There are two separate issues, as discussed by email last week:

  • the beam energies are the same, as mentioned by Nicolas
  • the combination of BES with ISR (PDF::Lepton = on) leads to other oddities, on the cross-section, and on the kinematics, resulting in a weird pz-imbalance of the hard event (post ISR) when sqrts is above the nominal energy.
    But the two issues do not look correlated.
    Cheers.
    E.
5

Thanks @eperez for confirming those are a priori no correlated.
This is not the topic of this post, but have you tried to redo the pz-imbalance with: BES and PDF:lepton=off? and are you saying that PartonLevel:ISR and ````PDF:lepton``` are the same thing (from the documentation it is not clear?

6

have you tried to redo the pz-imbalance with: BES and PDF:lepton=off ?

Yes I had checked this. The pz-imbalance is zero in this case (no pz-imbalance from the BES, and no imbalance from ISR)

and are you saying that PartonLevel:ISR and ````PDF:lepton``` are the same

No. PDF:Lepton does the ISRs in the equivalent photon approximation. With PDF:Lepton = on and PartonLevel:ISR = off, your ISR photons have zero pT. With PartonLevel:ISR = on in addition to PDF:Lepton = on, you have ISR photons with non-zero pT.

7

Hello,

Thanks a lot @eperez for your answers ! I will have a look at the Whizard sample at some point.

Regarding Pythia configs, just to make it clear I confirm that:

  • Setting BES on completely changes the cross-section, to some much too low value. As soon as it is off, the cross-section reported by Pythia seems fine.
  • PDF:lepton indeed switches the ‘0-pT’ ISR, and setting it off increases the ZH cross-section (as all e+e- interactions do happen at 240GeV precisely).

So the ‘right’ cross-section to use is the one computed with PDF:lepton=on (default), and BES=off.

8

So the ‘right’ cross-section to use is the one computed with PDF:lepton=on (default), and BES=off.

Yes. And this is indeed how the cross-sections that are provided (1) with the FCC-ee samples have been determined.
(1) the ones that you can read in the web page, or in the json dictionary file.