Discussion 10-21-24

Speakers and participants, please review the guidance provided on the main page. This agenda page is editable by anyone that has a Jefferson Lab computing account. Feel free to log in and post comments, questions, or answers to questions in the section below.

Meeting Location

The 22 GeV Open Discussions will be held here:

• Date/Time: Monday, October 21 at 12:30 PM Jefferson Lab Local Time
• Physical Location: CEBAF Center F224/5
• Virtual Location: Zoom Meeting Number 161 111 8017 (The password is the two-digit number that appears before "GeV" in the first sentence of this section.)

Agenda

• Measurement of the QCD coupling α_S with JLab@22 GeV (Alexandre Deur)

Comments and Questions

• (Add content here)

Minutes/Notes (D.S. Carman)

Local participants at JLab: 7; Remote participants: 36

Goal – To improve our knowledge of the strong coupling constant α_SS through a measurement program that will reduce the current uncertainties.

Presentation points:

• α_S is the most important quantity of QCD and is a key parameter of the Standard Model. However, it is by far the least known fundamental constant.
• There are a number of ongoing efforts to reduce the uncertainties of α_S, but there is no golden experiment to determine α_S precisely. The current strategy has been to combine independent measurements with larger uncertainties. Currently the best experimental determinations are at the Δα_S/α_S ~ 10^-2 level.
• There are good prospects of measuring α_S with JLab@22 GeV using an approach based on the Bjorken sum rule (measuring the integral over x of Γ₁^p-n vs. Q²). In this approach there is no need for an absolute cross section measurement. Only the determination of the Q² dependence is necessary with tight control over the point-to-point systematic uncertainties.
• The best measurement is one that spans the broadest range of Q² and x. The best prospect of extracting α_S is to combine data from JLab@22 GeV with the EIC. In this case it is estimated that Δα_S/α_S can be measured to the sub-percent level of 6.1x10^-3. Using the projections from the EIC measurement alone gives Δα_S/α_S ≳ 1.3%.
  • Such a combined JLab22 + EIC measurement can yield an uncertainty on α_S smaller than all of the available world data combined.
  • Measurements in different regimes of Q can be complementary. At low Q, measurements are sensitive to higher-order loop corrections. However, at higher Q, the measurements are essentially independent of these higher-order corrections and can be used to determine the value of α_S.

Questions/comments from discussion:

• Q: The biggest reduction in uncertainty on α_S seems to come from the combined JLab22+EIC data. What is the expected uncertainty from just the JLab22 measurement alone or from the JLab12+EIC measurement? Having this information available can help to better motivate the importance of the JLab22 program.
  • A: at the minimum Q² of EIC, data from JLab12 measure only 45% of the Bjorken integral, and this quickly decreases at larger Q². Hence, JLab12 would contribute only marginally to an α_S determination and thus cannot notably improve Δα_S/α_S. In contrast Jlab22 measures 90% of the integral, contributing dominantly.
• Q: Need a comparison of the Q² vs. x phase space for the different measurements to show overlap and differences.
  • A: This comment was made in connection with the question about the sophistication of the assessment that, with JLab22, one can reach Δα_S/α_S ~0.6%. The assessment is not as precise as that of EIC, which was done with extensive simulations and considerations. (The result of this study is published: Kutz et al. PRD 110 (2024) 7, 074004). While the estimate for JLab22 is not as sophisticated, there is also less need for such sophistication. The crucial questions are:
    • i) What are the actual (i.e., kinematics+detector efficiency) low-x limits and Q²? coverage of JLab22?
    • ii) What are the statistics?
    • iii) What are the experimental systematics?

The answers to (ii) and (iii) are straightforward: the statistics are always plenty for the Bjorken integral because data are gathered concurrently with other reactions, e.g., (semi-) exclusive ones for GPDs or transversity, which are more statistics-hungry. The systematics are already well-known because the same types of experiments were done at 6 and 12 GeV. Answering accurately question (1) is not crucial because wherever the low-x coverage will stop, EIC data will pick it up. We only need to have a reasonable answer (will JLab22 be missing 10% at low-x, or 40%?), which is straightforward. The minimum Q² reached is irrelevant because it will clearly be below 1 GeV² and for DIS, we consider only data at Q² >1 GeV². What is precisely the higher Q² limit is also not crucial because (1) EIC data will be available and (2) the sensitivity to α_S is reduced at large Q². Thus, a detailed Monte-Carlo simulation is not required at this stage. On the other hand, it was needed for EIC, where all the answers of the above questions were not as straightforward.

• Q: Make sure tables of data show all available results not just the PDG results. It was mentioned that there is a new precision result from LHC/ATLAS.
  • A: Noted.