Study of Longitudinal Event-plane Decorrelation in Heavy-ion Collisions with a Multiphase Transport Model

Measurements of pseudorapidity ($\eta$) dependence of local event-plane (EP) correlations provide novel inputs to the initial conditions in heavy-ion collisions. A factorization ratio of elliptic flow ($v_2$), $r_2(\eta) = \langle v_2(-\eta) \cos[2(\Psi_{\rm m,-\eta}-\Psi_{\rm f})]\rangle / \langle v_2(\eta) \cos[2(\Psi_{\rm m,\eta}-\Psi_{\rm f})]\rangle$, has been extensively used to quantify the decorrelation between two EPs at midrapidities ($\Psi_{{\rm m},\pm\eta}$) using a forward EP ($\Psi_{\rm f}$) as a reference. However, nonflow effects could cause features in $r_2$ similar to decorrelation. In this study, we shall employ a multiphase transport model (AMPT) to evaluate the sensitivity of several methods to both decorrelation and nonflow. Differential measurements of $r_2(\eta)$ and its slope, $F_2$, as well as their modified forms indicated possible nonflow effect contribution. We propose an alternative observable, $T_2 = \langle\sin(2(\Psi_{f} -\Psi_{m,1})) \sin(2(\Psi_{b} -\Psi_{m,2}))\rangle$, which has different sensitivity to dynamics of decorrelation from twist of initial participant matter and/or nonflow effect. We employ event-shape selection analysis technique to investigate features in $F_2$ and $T_2$, and discuss their physics implications.