Comments

[jtenavidal]

abstract: line 11: change to: radioactive effects must be accounted for in the simulation or corrected for in the data line 18: I think the abstract can be more general and not committing to a single approach. But rather write here that the model itself can be changed and the user should take into account the experimental setup in which the compared data was taken from.

Introduction: The introduction is great! Additional issues that should come up in the introduction: is the need iin external and intternal radiation for both electron and neutrino experiments In the future this methodology will be useful to add radiative effects for nuA interaction. The implementation depends on the experimental setup, One needs to account for the target width, photon and electron detection resolution and threshold. A user should modify the code using a config file and all needed parameters should be listed. There are cases in which this method cannot be used? we should mention them here

lines 22-82: The first section of the introduction is very long. I suggest splitting it to a few paragraphs and change its ordering: DUNE needs (motivation) similarity between electron and neutrino Current modelling and generators including their electron mode e4nu, allowing parallel analyses line 32: change "nuclear theory" to "classical nuclear theory"? to make sure the reader knows its basic line 32: change "eA data" to "eA inclusive data" line 37: The availability of eA exclusive data is a opening a new avenue and complements .... should be emphasise that this is a new approach, and because it's exclusive we cannot rely on the the known method of radiative corrections for inclusive data. you wrote it in lines 52-53, but it should be emphsised before. line 49: NRES should be explained line 58: exclusive electron scattering data line 60: electron/lepton beam instead of neutrino beam line 71: and perform parallel analyses to neutrino experiments after line 81: add plans from complementary efforts based in the table of the NuSTEC electron scattering review. lines 90-103: These paragraph reviews the preivous method and should include (1) method fr correcting inclusive measurement (2) Schwinger+Mo&Tsai you wrote about and mention that it for ep scattering. (3) current methods, JLab is sadlly not the most up to date, and there are many more it also has seperate packages: one for inclusive and others for simple exclusive. (4) why a new methdo is needed. you wrote it beautifully in lines 99-103 lines 104-116: The description of electron mode of generators will be better placed after explaining generators and thee current modelling. It shouold also mentiono that the electron mode is unified in GENIE, using the same code for nuclear model, FSI and the vector part of the interaction when possible. lines 117-128: this is a great paragraph and a fantastic way to finish the introduction. you can also mention the need to facilitate many optional models and take into account the experimental setup conditinos.

section II methodology is very well written! An extra line mentioning that in the future radiative effects for nuA interaction will be implemented by omitting step1 and modifying step5. section III line 188: it's elastic and not quasielastic if it's on H we need to decide if we're repeating the procedure or referring to the SIMC writeup. We don't do the same for the non-QE case. I'm not sure what is the right way to go. We should consult with the rest. In the original implementation if the photon momentum was below Emin the electron was not radiated. Is it still the case? If so, should be mentioned. figure 3 caption: the caption is unclear related to the two samples, I would elaborate on what is the pink dot.

line 320: typo "A event graph" to "An event graph"

The rest of the paper is honestly very good, I didn't have any more suggestions for this round

[jtenavidal]

Dear Jula Thank you for all the work that you have done so far on producing GENIE cross sections for 1 and 4 GeV electron beams.

I have divided your  1 and 4 GeV cross sections by 225,000.  So till we figure out your normalization, we are unable to

determine if the GENIE normalization is correct

I enclose comparisons to two electron scattering measurements, . I am wondering about the GENIE parameters. I thought SuSv2 uses Psi scaling, so I do not understand the ground state local Fermi Gas for QE scattering, so I assume that Local Fermi Gas is just for Resonance and DIS. Is that true?

In any case, the comparisons below at 1 GeV 40 deg AND 4 GEV 15 deg for Q^2 _QE. = 1.108 AND 0.97 GEV, respectively. I am unable to do RL RT separation till I get the complete set of energies.

However, so far it appears that the QE peak is shifted to lower \nu by about 40 MeV. This is a binding energy/optical potential modeling Issue (which is addressed in my paper with my student Tejin Cai).

It also appears that the Delta cross section is too low relative to QE, which is Berger-Sehgal problem. If these issues are also there at lower energies it is a serious problem. But will find out

For neutrino scattering Lower Q^2 is more relevant . So your results at lower energies will be of more interest to neutrino experiments. For those energies we can do a direct comparison to Barreau cross section data. Then our RL RT analysis of the GENIE cross section will reveal the modeling of longitudinal versus the modeling of the transverse contribution.s

  We will post our RL RT paper this month (but with only comparisons to nuclear theory).  We would like to write another
Paper which is comparison to MC generators, which you will be a co-author for the  GENIE comparisons and Artur Ankowski will
be an author of the NuWRo comparisons (which are all done).  But NuWRo is only QE,  and only the neutrino mode has MEC.
So their transverse cross sections in electron mode a too low.  The longitudinal cross sections are too high because
they do not have longitudinal quenching (as revealed by our RL RT analysis.

I am trying to get the ACHILLES cross sections,, but I have not heard from them yet. So if they do not respond quickly, we will just do NuWRo and GENIE comparisons for C12 and then move to Ca40.

Best regards - Arie

[jtenavidal]

Dear Julia - actually the 1 GeV comparison is at Q2= 0.41 GeV (not 1.18). So we have two cross sections at Q2=0.41 GeV2 and at Q2=0.97 GeV2. The Delta_1238 cross section appears underestimated (assuming the QE is correct),

It looks like \nu(QE) may be underestimated by 37 MeV. See comparison below (first plot) before and after adding 37 MeV to \nu.

However, I am comparing your 1 GeV 40 degrees to the measurement by Sealock  at 1.08 GeV 37.5 degrees

so it is not an exact comparison.

 Also for  the 4 GeV 15 degree data the normalization is important and the energy  shift depends on normalization and on the contributions from the Delta.
So again it is not clear till we get the absolute normalization,

So I would wait to see how we compare to the Barreau data (the list of lower energies that I send you for which we will have exact comparisons).

The comparison of previous versions of GENIE to 3 of the Barreau cross section for C12 (0.24 and 0.56 GeV) Is given in
arXiv:2009.07228v2 [nucl-th] 13 Apr 2021

It shows GENIE in better agreement. (See second plot attached).

In any case, in the Barreau comparisons  we can compare both the cross sections at a large range of
energies (there are many of them) and the extracted    RL RT, which differentiate between longitudinal and transverse modeling over a very large range Q2 and nu.

• Best regards Arie

[jtenavidal]

Yep, nothing to worry about here!

So this confusion is added because of the facorization we do in the generator. There's the inclusive cross-section that give xsecs as a function of Enu and lepton kinematics that's made using some nuclear model and is integrated over the hadronic system. For SuSAv2-QE, this is RMF but in an effective sort of way (although the formalism is from RFG). But this is where SuSAv2 stops, it just gives lepton kinematics and nothing else. In this sense, no matter what nuclear model you choose in GENIE, the lepton kinematics in the SuSAv2 model will not change and are effectively from an RMF nuclear model (and not whatever nuclear model the user chooses). (Note that in other models the choice of nuclear models does impact the outgoing lepton kinematics because the factorization that's used is a little different).

But for the full GENIE implementation we need to predict hadron kinematics too, and here we do some ad-hoc mess in order to guess outgoing nucleon kinematics event-by-event from the SuSAv2 model. For this bit, whatever we do is wrong, but we may as well use the "best" model GENIE has and LFG is certainly better than RFG (we do say we do this in the implementation paper, but it's easy to miss).

Overall, the nuclear model you use in GENIE shouldn't matter at all whilst you care about lepton kinematics and any choice is more or less as wrong as any other for outgoing nucleon kinematics.

There's some discussion of this issue here:

https://arxiv.org/abs/2302.12182

And then this approach is compared to a proper exclusive calculation both in the implementation paper you linked (this is what Figs. 4 and 12 are about) and then is further explored in these papers that contain the SuSAv2 calculations alongside the equivalent "ED-RMF" exclusive calculation.

https://arxiv.org/abs/2304.01916 https://arxiv.org/abs/2207.02086

Let me know if this isn't clear!

Best,

Stephen