Correcting for cutoff dependence in backward evolution of QCD parton showers

TL;DR

This paper investigates the impact of cutoff-dependent terms in the backward evolution of QCD parton showers, revealing limitations in reproducing cutoff-independent PDFs and proposing corrections for improved phenomenological accuracy.

Contribution

It computes cutoff-dependent no-emission probability corrections and proposes a consistent approach using cutoff-dependent PDFs and matrix element corrections.

Findings

Cutoff-dependent terms significantly affect backward shower evolution.

Reproducing cutoff-independent PDFs with backward showers is fundamentally limited.

Computed first-order corrections for Drell-Yan and Higgs production processes.

Abstract

Monte Carlo event generators for hard hadronic collisions depend on the evolution of parton showers backwards from a high-scale subprocess to the hadronization scale. The evolution is treated as a branching process with a sequence of resolvable parton emissions. The criterion of resolvability involves cutoffs that determine the no-emission probability (NEP) for a given range of the evolution scale. Existing event generators neglect cutoff-dependent terms in the NEP that, although formally power-suppressed, can have significant phenomenological effects. We compute such terms and study their consequences. One important result is that it is not possible for the backward shower to faithfully reproduce the cutoff-independent parton distribution functions (PDFs) used to generate it. We show that the computed NEP corrections mitigate but do not eliminate this problem. An alternative approach is to use cutoff-dependent PDFs that are consistent with the uncorrected NEP. Then one must apply cutoff-dependent corrections to hard subprocess matrix elements. We compute those corrections to the first nontrivial order for the Drell-Yan process and for Higgs production by gluon fusion.