Viscous shock solutions to the stochastic Burgers equation (Q2065718)

Consider the stochastic Burgers equation for \(u: \mathbb{R}\times\mathbb{R}\to\mathbb{R}\): \[ d u(t,x)= \frac 1 2 [\partial_x^2 u(t,x)-\partial_x(u^2)(t,x)] d t + d (\partial \{\rho * W\})(t,x), \] where \(\rho\in C^\infty(\mathbb{R})\cap H^1(\mathbb{R})\), and \(W\) is the cylindrical Brownian motion on \(L^2(\mathbb{R})\). This paper introduces a new notion of viscous shock solution which connects ``top'' and ``bottom'' spatially stationary solutions of the same equation. It is shown that such shocks admit time-invariant measures when viewed in their own reference frames. Under such a measure, the viscous shock is a deterministic function of the bottom and top solutions and the shock location, while the measure of the bottom and top solutions must be tilted to account for the change of reference frame. A convergence result to these stationary shock solutions is derived for solutions initially connecting two constants.