Low-energy $e^+\,e^-\toγ\,γ$ at NNLO in QED
Tim Engel, Marco Rocco, Adrian Signer, Yannick Ulrich
TL;DR
This work delivers a fully differential NNLO QED calculation of $e^+\,e^- \to \gamma\gamma$ at low energies, implemented in the McMule framework to provide precise predictions for luminosity-sensitive observables in $e^+e^-$ colliders up to a few GeV. The calculation separates photonic and fermion-loop contributions, employing massification for double virtual terms, OpenLoops for real-virtual corrections, and a dispersive approach for vacuum polarisation, while neglecting subdominant heavy-lepton and hadronic LbL effects. The results show NNLO photonic corrections at the permille level with non-photonic effects generally smaller, validating comparison with NLO+PS predictions and supporting a total theoretical accuracy around $0.1\%$ for typical low-energy scenarios. The study includes differential distributions for KLOE- and Belle-like setups, illustrating the impact of NNLO corrections on angular observables and reinforcing the utility of McMule for precision luminosity determinations at low-energy colliders.
Abstract
We present a fully differential computation of $e^+\,e^-\toγ\,γ$ at next-to-next-to-leading order in QED. The process has been implemented into McMule, completing its set of next-to-next-to-leading-order calculations for the most important $2 \to 2$ processes. The results allow for generic applications to electron-positron colliders with centre-of-mass energies up to a few GeV, particularly for luminosity measurements.
