This paper presents a practical way to predict the production of Drell–Yan lepton pairs in proton collisions from very low to very high transverse momentum. The authors build a particle-level prediction that combines very-high-order calculations, a non-perturbative input from lattice QCD, and a statistical reweighting step so the results can be used inside standard Monte Carlo simulations with positive event weights.

What the researchers did: they put together several state-of-the-art ingredients. At small to moderate transverse momentum they use high-order resummation and fixed-order results (the abstract reports N3LO fixed-order and approximate N4LL′ resummation). For high transverse momentum they include tree-level predictions with up to four, five, and six extra jets and approximate next-to-leading-order (NLO) electroweak corrections. A key non-perturbative ingredient, the Collins–Soper kernel, is taken from continuum-limit lattice QCD calculations. The Collins–Soper kernel controls part of the rapidity evolution that affects the low-momentum region and was previously fitted to data; using lattice results makes that part of the prediction ab initio.

How it works at a high level: the computation uses the factorization picture from soft-collinear effective theory. Short-distance effects are handled by fixed-order hard functions. Radiation that is soft or collinear to the beam is described by beam and soft functions and organized by resummation. Where perturbation theory breaks down at large impact parameter (a mathematical variable used in the resummation), the lattice-determined Collins–Soper kernel supplies the missing non-perturbative input. The authors add a small, empirically tuned Gaussian parameter that controls beam-function corrections; this single width parameter is fitted to low-transverse-momentum ATLAS data at 8 and 13 TeV.