Prof. Gavin Salam FRS

Abstract:

We formulate a collinear partonic shower algorithm that achieves next-to-single-logarithmic (NSL, αsnLn−1) accuracy for collinear-sensitive non-singlet fragmentation observables. This entails the development of an algorithm for nesting triple-collinear splitting functions. It also involves the inclusion of the one-loop double-collinear corrections, through a z-dependent NLO-accurate effective 1 → 2 branching probability, using a formula that can be applied more generally also to future full showers with 1 → 3 splitting kernels. The specific NLO branching probability is calculated in two ways, one based on slicing, the other using a subtraction approach based on recent analytical calculations. We close with demonstrations of the shower’s accuracy for non-singlet partonic fragmentation functions and the energy spectrum of small-R quark jets. This work represents an important conceptual step towards general NNLL accuracy in parton showers.

Abstract:

We report on a major milestone in the construction of logarithmically accurate final-state parton showers, achieving next-to-next-to-leading-logarithmic (NNLL) accuracy for the wide class of observables known as event shapes. The key to this advance lies in the identification of the relation between critical NNLL analytic resummation ingredients and their parton-shower counterparts. Our analytic discussion is supplemented with numerical tests of the logarithmic accuracy of three shower variants for more than a dozen distinct event-shape observables in Z→qq[over ¯] and Higgs→gg decays. The NNLL terms are phenomenologically sizeable, as illustrated in comparisons to data.