Disentangling flow and nonflow correlations via Bayesian unfolding of the event-by-event distributions of harmonic coefficients in ultrarelativistic heavy-ion collisions

Abstract

The performance of the Bayesian unfolding method in extracting the event-by-event (EbyE) distributions of harmonic flow coefficients v_n is investigated using a toy model simulation, as well as simulations based on the HIJING and AMPT models. The unfolding method is shown to recover the input v_2-v_4 distributions for multiplicities similar to those observed in Pb+Pb collisions at the LHC. The effects of the nonflow are evaluated using HIJING simulation with and without a flow afterburner. The probability distribution of v_n resulting only from nonflow in HIJING is nearly a Gaussian and can be largely suppressed in the data-driven unfolding method used by the ATLAS Collaboration. The residual nonflow effects have no appreciable impact on the v_3 distributions, but are found to affect the tails of the v_2 and v_4 distributions; these effects manifest as a small simultaneous change in the mean and standard deviation of the $v_n$ distributions. For the AMPT model, which contains both flow fluctuations and nonflow effects, the reduced shape of the extracted v_n distributions is found to be independent of pT in the low pT region, similar to what is observed in the ATLAS data. The prospect of using the EbyE distribution of the harmonic spectrum aided by the unfolding technique as a general tool to study azimuthal correlations in high energy collisions is also discussed.