Investigating the collision energy dependence of η/s in RHIC beam energy scan using Bayesian statistics

We determine the probability distributions of shear viscosity over the entropy density ratio $\eta/s$ in Au+Au collisions at $\sqrt{s_{NN}}=19.6, 39$, and $62.4$ GeV, using Bayesian inference and Gaussian process emulators for a model-to-data statistical analysis that probes the full input parameter space of a transport+viscous hydrodynamics hybrid model. We find the most likely value of $\eta/s$ to be larger at smaller $\sqrt{s_{NN}}$, although the uncertainties still allow for a constant value between 0.10 and 0.15 for the investigated collision energy range.