Determining fundamental properties of matter created in ultrarelativistic heavy-ion collisions

Abstract

Posterior distributions for physical parameters describing relativistic heavy-ion collisions, such as the viscosity of the quark-gluon plasma, are extracted through a comparison of hydrodynamicbased transport models to experimental results from 100A GeV + 100A GeV Au+Au collisions at the Relativistic Heavy Ion Collider (RHIC). By simultaneously varying six parameters and by evaluating several classes of observables, we are able to explore the complex intertwined dependencies of observables on model parameters. The methods proved a full multi-dimensional posterior distribution for the model output, including a range of acceptable values for each parameter, and reveal correlations between them. The breadth of observables and the number of parameters considered here go beyond previous studies in this field. The statistical tools, which are based upon Gaussian Process emulators, are tested in detail and should be extendable to larger data sets and a higher number of parameters.

Toward a deeper understanding of how experiments constrain the underlying physics of heavy-ion collisions

“…The details of the model are expounded in more detail in [3,4], but we will briefly review the basics and focus on describing the 14 parameters varied in this analysis along with the set of observables. Our model consists of the generation of an initial state which is then fed into a 2-dimensional hydrodynamics simulation followed by transition to a microscopic simulation, known as a hadronic cascade, at a transition temperature of T 0 = 165 MeV.…”

“…The other five described the initial state for Pb+Pb collisions at 1.38A TeV + 1.38A TeV from the LHC (Large Hadron Collider at CERN). The initial transverse energy density profile that instantiates the hydrodynamics has the form [3] (x, y) = f wn wn (x, y)…”

“…In the next two sections we review the model, data and techniques used to generate our previous results in [3,4], presenting a wider range of results from the projections of the MCMC procedure than were presented in [4]. In Sec.…”

“…In contrast, an MCMC exploration of parameter space provides not only the uncertainties for each parameter, but the covariances with other parameters along with the complete shape of the posterior likelihood. The methods of [3] indeed provide the complete likelihood distribution, and are now extracting the field's first rigorous quantitative conclusions from heavy-ion collisions at the highest energies.…”

“…Recent progress in methodology has overcome this difficulty through the use of Gaussian process emulators which can be trained to accurately reproduce the results of actual simulated models. The reduced computational load has opened the door to quantitative analyses of model parameters which has allowed for systematic constraints of the nuclear equation of state and of the QGP shear viscosity [3,4] and details of the initial state [5].…”