An extraction of the Collins–Soper kernel from a joint analysis of experimental and lattice data Preprint: DESY-25-145; FERMILAB-PUB-25-0784-T; INT-PUB-25-026; MIT-CTP/5946

Artur Avkhadiev, Valerio Bertone, Chiara Bissolotti, Matteo Cerutti, Yang Fu, Simone Rodini, Phiala Shanahan, Michael Wagman, Yong Zhao · 2025

Read An extraction of the Collins–Soper kernel from a joint analysis of experimental and lattice data Preprint: DESY-25-145; FERMILAB-PUB-25-0784-T; INT-PUB-25-026; MIT-CTP/5946 by Artur Avkhadiev, Valerio Bertone, Chiara Bissolotti, Matteo Cerutti, Yang Fu, Simone Rodini, Phiala Shanahan, Michael Wagman, Yong Zhao (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-26489.txt.

Introduction

Artur Avkhadiev, with eight co-authors, pins down a key piece of the strong force's behaviour in moving particles. The piece is the Collins-Soper kernel, a function that guides how parton patterns shift with energy in transverse-momentum-dependent distributions. Those distributions describe how a proton's parts spread in sideways motion as well as along its path.

Avkhadiev and the team combine lab data with lattice QCD, the grid-based form of the strong force theory, in a single joint fit. A neural network shapes the unknown functions, and a Bayesian reweighting step folds the lattice results in. Adding lattice input moved the kernel's central value by about ten percent. It also cut the uncertainty by forty to fifty percent. The work shows how grid calculations can sharpen fits built on lab data alone.

Catalog record

CREATORArtur Avkhadiev, Valerio Bertone, Chiara Bissolotti, Matteo Cerutti, Yang Fu, Simone Rodini, Phiala Shanahan, Michael Wagman, Yong Zhao
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE19 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2510-26489

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

An extraction of the Collins–Soper kernel from a joint analysis of experimental and lattice data Preprint: DESY-25-145; FERMILAB-PUB-25-0784-T; INT-PUB-25-026; MIT-CTP/5946 Artur Avkhadiev, Valerio Bertone, Chiara Bissolotti, Matteo Cerutti, Yang Fu, Simone Rodini, Phiala Shanahan, Michael Wagman, Yong Zhao Abstract We present a first joint extraction of the Collins–Soper kernel (CSK) combining experimental and lattice QCD data in the context of an analysis of transverse-momentum-dependent distributions (TMDs). Based on a neural-network parametrization, we perform a Bayesian reweighting of an existing fits of TMDs using lattice data, as well as a joint TMD fit to lattice and experimental data. We consistently find that the inclusion of lattice information shifts the central value of the CSK by approximately 10% and reduces its uncertainty by 40-50%, highlighting the potential of lattice inputs to improve TMD extractions. I Introduction This work presents the first effort to incorporate lattice QCD data into a phenomenological extraction of transverse-momentum-dependent distributions (TMDs), to better constrain the Collins-Soper kernel (CSK). This shares the same idea as past combined analyses of lattice and experimental data towards a better understanding of the hadronic structure through the extraction of collinear distributions. These include analyses of unpolarized, longitudinally-polarized, and transversity parton distribution functions (PDFs) Lin et al. 2018 ; Cichy et al. 2019 ; Del Debbio et al. 2020 ; Del Debbio et al. 2021 ; Bringewatt et al. 2021 ; Barry et al. 2022 ; Hou et al. 2023 ; Karpie et al. 2024 ; Ablat et al. 2024 ; Cocuzza et al. 2025 ; Ablat et al. 2025 ; Good et al. 2025 ; Barry et al. 2025a , and generalized parton distributions Guo et al. 2023 ; Cichy et al. 2024 ; Guo et al. 2025 . However, this is the first time that a similar strategy is employed in the context of a TMD extraction (see also Ref. Cridge et al. 2025 for a similar effort in this direction). In this context, the CSK plays a special role in that it encodes information on the QCD vacuum rather than on the actual structure of hadrons. In fact, there is no experimental observable directly sensitive to CSK, so lattice is the only source of direct information. Improved determinations of the CSK are important to further sharpen control and reduce model-dependence in TMD evolution, enhance the precision of QCD predictions of multiple processes, and test the validity of factorization in the nonperturbative regime. Formally, the CSK K is the expectation value of a Wilson loop running along the light-front directions and characterized by a transverse displacement \mathbf{b}_{\rm T} . When b\equiv|\mathbf{b}_{\rm T}| is small enough, the CSK can be computed in perturbation theory. As customary, radiative corrections introduce ultraviolet divergences, which, upon renormalization, lead to the introduction of the renormalization scale \mu . The dependence of K on \mu is governed by the renormalization-group equation: \frac{\partial K(b,\mu)}{\partial\ln\mu}=-\gamma_{K}(\alpha_{s}(\mu))\,, (1) where \gamma_{K} , a.k.a. cusp anomalous dimension, is computable in perturbation theory and is currently fully known to four loops Moult et al. 2022 ; Duhr et al. 2022 . Introducing the scale \mu_{b}=2e^{-\gamma_{\rm E}}/b , where \gamma_{\rm E} is the Euler-Mascheroni constant, the solution to Eq. ( 1 ) reads: K(b,\mu)=K(b,\mu_{b})-\int_{\mu_{b}}^{\mu}\frac{d\mu^{\prime}}{\mu^{\prime}}\gamma_{K}(\alpha_{s}(\mu^{\prime}))\,, (2) where K(b,\mu_{b}) is also known to four loops Moult et al. 2022 ; Duhr et al. 2022 . When b grows, such that \mu_{b}\lesssim\Lambda_{\rm QCD} , nonperturbative effects dominate and Eq. ( 2 ) becomes unreliable. Nonperturbative affects are accounted for by modifying Eq. ( 2 ) as follows: K(b,\mu)=K(b_{*},\mu_{b_{*}})-\int_{\mu_{b_{*}}}^{\mu}\frac{d\mu^{\prime}}{\mu^{\prime}}\gamma_{K}(\alpha_{s}(\mu^{\prime}))-g_{K}(b)\,, (3) with \mu_{b_{*}}=2e^{-\gamma_{\rm E}}/b_{*} . Here, b_{*}\equiv b_{*}(b) behaves linearly for small values of b ( b_{*}\sim b ) and saturates to some b_{\rm max}\ll\Lambda_{\rm QCD}^{-1} for large b . The purpose of b_{*} is to guarantee that the first two terms in the r.h.s. of Eq. ( 3 ) are evaluated in the perturbative regime where \alpha_{s}\ll 1 . The role of the functions g_{K} is to encode nonperturbative effects that arise from the large- b region. A typical parametrization is: g_{K}(b)=2g_{2}^{2}b^{2}\,, (4) with the parameter g_{2} to be determined from observables sensitive to the CSK. Other models for g_{K} , with at most two free parameters, have been used in recent TMD fits (see e.g. Refs. Moos et al. 2025 ; Aslan et al. 2024 ; Bacchetta et al. 2020 ). However, in this context, we found no evidence for the need of a more complicated functional form. The purpose of this work is therefore to determine the parameter g_{2} from a joint analysis of experimental and lattice data. II Baseline TMD fit A determination of g_{2} from experimental data implies an extraction of TMDs. TMD analyses have achieved a remarkable level of sophistication: they are often based on broad, multiprocess data sets, they achieve high perturbative accuracy, and employ advanced methodological frameworks (see, e.g., Refs. Bacchetta et al. 2024 ; Bacchetta et al. 2022 ; Moos et al. 2024 ; Moos et al. 2025 ; Camarda et al. 2024 ; Barry et al. 2025b ; Cuerpo et al. 2025 ; Kang et al. 2024 ; de Florian and Grazzini 2005 ; Camarda et al. 2025 ). In this work, we rely on the TMD extraction presented in Ref. Bacchetta et al. 2025 . In that analysis, TMD parton distribution functions (PDFs) were extracted from a comprehensive set of Drell-Yan data using next-to-next-to-next-to-leading logarithmic (N 3 LL) accurate predictions. The nonperturbative part of TMD PDFs, which also involves g_{2} , was parametrized through a neural network (NN). The flexibility of the NN enabled a solid estimate of the nonperturbative parameters and their uncertainties. That analysis obtained: g_{2}^{\rm Baseline}=0.186\pm 0.033\,. (5) In this Letter, we will use the numerical framework and the determination of Ref. Bacchetta et al. 2025 as a baseline to quantify the impact on g_{2} of lattice data for the CSK in the context of a TMD phenomenological extraction. III Lattice data Over the last five years, several efforts have been made to extract the CSK from lattice QCD following the original proposals in Refs. Ji et al. 2015 ; Ebert et al. 2019a ; Ebert et al. 2019b ; Ji et al. 2020a ; Ji et al. 2020b ; Ji and Liu 2022 and follow-up works in Refs. Vladimirov and Schäfer 2020 ; Rodini and Vladimirov 2023 ; Deng et al. 2022 ; Zhao 2024 . Exploratory calculations based on quasi-TMD beam functions have been performed in quenched Shanahan et al. 2020 and dynamical Shanahan et al. 2021 ; Shu et al. 2023 lattice QCD. Analogous calculations based on quasi-TMD wave functions Ji et al. 2020a ; Ji and Liu 2022 were also carried out Zhang et al. 2020 ; Chu et al. 2022 ; Li et al. 2022 ; Schlemmer et al. 2021 ; Shu et al. 2023 ; Alexandrou et al. 2025 . Recent lattice QCD calculations of the quark CSK using quasi-TMDs in Coulomb gauge Bollweg et al. 2024 ; Bollweg et al. 2025 have also shown promise for significant reduction in statistical uncertainty. However, at present there is only a single lattice QCD determination of the CSK with full systematic control, including, for the first time, the continuum a\to 0 extrapolation Avkhadiev et al. 2023 ; Avkhadiev et al. 2024 . The

Read the complete text