Two-particle cumulant distribution: a probe of "true" elliptic flow
Satya Ranjan Nayak, Akash Das, B. K. Singh
TL;DR
This paper addresses the challenge of distinguishing true elliptic flow from non-flow in small systems, specifically d-Au collisions at $\sqrt{s_{NN}}=200$ GeV. It uses Angantyr (a PYTHIA8 extension) to model non-flow contributions to the two-particle cumulant $c_2$ and contrasts these with HYDJET++-modeled true flow, analyzing event-by-event distributions across varying $\Delta\eta$ windows and $N_{ch}$. The key finding is that non-flow distributions are non-Gaussian, with high skewness and kurtosis that grow with $\Delta\eta$ and multiplicity, while true-flow distributions are Gaussian-like. The study proposes using the skewness and kurtosis of the $c_2$ distribution as a light, cross-check diagnostic to separate non-flow from true hydrodynamic flow, potentially improving $v_2$ extractions in small systems.
Abstract
In this work, we have shown the two-particle correlations of charged hadrons in d-Au collisions at 200 GeV. These correlations were studied at different multiplicities and pseudorapidity intervals. The two-particle correlations arise due to Color Reconnections, resonance decays, jet correlations, and hadronic rescattering. These correlations are inversely proportional to multiplicity but remain unaffected for larger pseudorapidity windows. We treated these correlations as distributions and calculated their skewness and kurtosis. The non-flow distributions deviate greatly from a Gaussian distribution and have high skewness and kurtosis. The "true" elliptic flow distributions resemble Gaussian distributions; they have significantly lower skewness and kurtosis. We suggest that if the two-particle cumulant flow is treated as an event-by-event distribution, its skewness and kurtosis can be instrumental in distinguishing true flow and non-flow.