Anomalous shock fluctuations in TASEP and last passage percolation models (Q2257117)

The totally asymmetric simple exclusion process (TASEP) is an interacting particle system on the integers \(\mathbb Z\). Particles jump at constant rate onto the nearest-neighbor site at the right, provided that it is unoccupied by a particle. The particle density at time \(t\) solves an approximation of Burger's partial differential equation, and the discontinuities of the solution are called shocks. Earlier works on TASEP models started from product measures, which are the only invariant measures. For initial product measures with a shock at the origin, fluctuations of the shock location are Gaussian of order \(t^{1/2}\) in the asymptotic limit of time. Exploiting the earlier established connection between last passage percolation (LLP) on the integer plane \(\mathbb{Z}^2\) and the TASEP, the authors derive a generic theorem, which dictates that ``the distribution function of a generic LLP is the product of two distribution functions corresponding to two simpler last passage problems.'' Their generic theorem is applied to a variety of LLP and TASEP models through the verification of the assumptions of the theorem. For a certain TASEP model that motivated their work, the model is started from a hybrid configuration \((\dots, 0, 1, 0, 1, \dots)\) interpreting 1 as occupied. The particles attempt to jump at two different rates depending on whether they are initially located at the left or the right of the origin. Shock fluctuations of order \(t^{1/3}\) are expressed in the long term as a product two Airy\({}_1\) processes, whose one-dimensional marginal distribution is described by the Tracy-Widom distribution of the largest eigenvalue in the Gaussian orthogonal ensemble of random matrices. This accumulates evidence for the universality of the Airy\({}_1\) process, for which there is a certain lack of findings. In other corollaries of the generic theorem, the Airy\({}_2\) process is involved, whose one-dimensional marginal distribution is described by the Tracy-Widom distribution of the largest eigenvalue in the Gaussian unitary ensemble.