SciPost Submission Page
Nonperturbative effects in triple-differential dijet and Z+jet production at the LHC
by Stefan Gieseke, Maximilian Horzela, Manjit Kaur, Dari Leonardi, Klaus Rabbertz, Aayushi Singla, Cedric Verstege
Submission summary
| Authors (as registered SciPost users): | Klaus Rabbertz |
| Submission information | |
|---|---|
| Preprint Link: | https://arxiv.org/abs/2412.19694v2 (pdf) |
| Date submitted: | Dec. 1, 2025, 9:43 p.m. |
| Submitted by: | Klaus Rabbertz |
| Submitted to: | SciPost Physics |
| Ontological classification | |
|---|---|
| Academic field: | Physics |
| Specialties: |
|
| Approaches: | Experimental, Phenomenological |
Abstract
In comparisons of precision measurements at colliders to the most accurate predictions available in perturbative quantum chromodynamics (QCD), it is required to correct for nonperturbative effects. By means of Monte Carlo event generators this article investigates the impact of such nonperturbative effects on two processes relevant for precision determinations of the strong coupling constant and the proton structure: triple-differential dijet and Z+jet production. We observe significant differences between the two processes. Whether this non-universal behaviour is realized in nature remains an open question. We therefore propose a triple-differential measurement of the underlying event in Z+jet production.
Author indications on fulfilling journal expectations
- Provide a novel and synergetic link between different research areas.
- Open a new pathway in an existing or a new research direction, with clear potential for multi-pronged follow-up work
- Detail a groundbreaking theoretical/experimental/computational discovery
- Present a breakthrough on a previously-identified and long-standing research stumbling block
Current status:
Reports on this Submission
Report
Aside from some more minor comments listed below, I have one primary concern with the manuscript: the analysis of the non-perturbative correction function for Z+jets. The authors observe that, in this process, the corrections are significantly larger than those obtained in dijet production. However, I would argue that their “LO” predictions are not well suited to study these effects, given that all observables measure quantities against the leading jet, which at LO is mostly produced via the shower. While this is a good approximation in certain phase-space regions, it certainly is not valid in all of them, and especially not in very hard regions where a full matrix element should be used. It is not entirely clear to me what exactly is included in the NLO simulation (see one of the more minor points below), but ideally the first jet should also be simulated at NLO in order to obtain a trustworthy prediction. Is it not clear to me if the NLO samples fullfill this, since producing two muons at NLO would have the first jet only at leading order. The merged Sherpa sample seems well suited and consistently shows the smallest corrections.
Furthermore, the observations are slightly obscured by fitting according to Eq. (7). Looking at the plots, it seems like only very few bins actually show significant deviations and I think it would be more insightful to show the corrections bin-by-bin for these regions.
However, if the findings hold up, I do not see how one should conclude that “NP corrections […] may not be entirely of NP origin.” Given the authors’ analysis, a more appropriate assessment would be that NP corrections obtained in this way depend on the perturbative input used in the simulation. I think this should be formulated more carefully in the conclusion.
Requested changes
-
Revise the analysis of the non-perturbative correction factor for Z+jets. I do not think this requires rerunning parts of the Monte Carlo, but I think the analysis should be performed based purely on the "NLO" results. If not, I think convincing arguments should be made, why the "LO" simulation provides sensible results for this setup.
-
I do not understand the reason for fitting according to Eq. (7). Looking at Fig. 3, the errors introduced seem small for most yy bins and do not seem to justify obscuring the plots by using a fit instead of showing a bin-wise correction factor. Is this typical experimental procedure? I assume that the error bars in Figs. 10–12 correspond to statistical variations, and they mostly seem small enough to me.
-
Similarly, Eq. (6) is later described as the “conventional” method to obtain NP corrections. It would be helpful to reference calculations and/or measurements that make use of this technique.
-
The generator setups are described relatively briefly and not in sufficient detail for reproduction. Given the impact of higher-order contributions, either via NLO or merging, it would be helpful to add more details on what went into the simulations, for example the merging cuts, matching scheme, ...
-
In multiple places, the authors refer to Z-boson (+jets), e.g. in Section 3.2. I am wondering whether the simulated samples are truly Z+jets samples, or whether photons are taken into account as well. I doubt that there is a large impact on the analysis given the phase-space cuts, but the formulation is potentially misleading in several places.
Recommendation
Ask for minor revision
Strengths
- Discusses non-perturbative (NP) corrections to important collider measurements in a clear and systematic way.
- Uncovers an interesting point about these NP corrections, that there are regions in which they are significant and appear to be non-universal.
Weaknesses
- Does not go as far in exploring possible explanations of the non-universal behaviour as one might hope (but this could be seen as a stimulation to the community to do so).
- Formulates the NP corrections as multiplicative rather than as a migration of jets up or down the pT distribution (which can be important since the two processes they compare have different pT slopes).
- Also formulates the NP corrections in a way that implies the parton shower of an event generator encapsulates the perurbative results and that the other components (hadronization and MPI) define the NP corrections.
Report
Requested changes
- There is no discussion of what the “Herwig LO”, “Herwig NLO”, “Sherpa LO” and “Sherpa NLO” results actually mean. Both programs have various options and running conditions and it is not at all clear just from these names which were used. Which of the parton shower algorithms, with which parameters, which MPI options and parameters, etc. Even the appellations “LO” and “NLO” are ambiguous, especially for the Z case - the LO process could be Z production or Z+jet production. And, in either case, what NLO matching algorithm/options are used? If the authors don’t want to clutter the main text with these details, an addition to the appendix would be ideal, but this really has to be spelled out in more detail for readers to make sense of the results.
- The paper takes a very traditional view of NP corrections as “everything that takes place in an event generator after the parton shower”. However, it is more commonly understood nowadays that the infrared cutoff of the parton shower is also the ultraviolet cutoff of the hadronization model and that one cannot separate them unambiguously - the cutoff dependence of the shower should be considered part of the NP correction. While I am not proposing that the authors change their approach at this stage, I do think it would be helpful to at least raise awareness of this issue, e.g. in the text after Eq. (6). At least to mention what the cutoff value is and how it is implemented (presumably as a pT cut) and remind readers that the parameters of the hadronization model are tuned in tandem with this value.
- It is common to think of NP corrections in terms of a migration (jets are shifted to higher or lower pt by radiation into or out of the cone), rather than a multiplicative rate correction. Again, I am not advocating a change of approach, but it would be helpful to mention this, and the reason for preferring a multiplicative correction, e.g. in the discussion around Eq. (7).
- In the discussion of section 5.1, there are a couple of small errors. It says that “In dijet or minimum bias events, the activity in the towards and away region is also strongly associated with the transverse momentum of the leading particle … The main driver of this is the parton radiation … “. Actually the main driver of the pT in the towards region is the momentum of the primary jet and in the away region the momentum of the other jet that recoils from it - the parton radiation mainly just serves to spread out this momentum, but the majority of it remains in the toward and away regions. And in the next paragraph, “The hardness of the event is associated with the transverse momentum of the Z boson” - this is somewhat true for the away region, which should contain the jet recoiling from the Z, but soft radiation into the transverse and toward region is determined by the scale of the hard process, which is something like the transverse mass - even Z events with zero pT produce soft radiation up to order mZ.
- Finally, a very very minor presentational thing - the page break in the middle of section 6 is very poor and, unless adding extra text to address the above points improves it anyway, I would suggest forcing the full-page figures to appear before starting the Summary and outlook section.
Recommendation
Publish (meets expectations and criteria for this Journal)