On stationary distributions for the KPP equation with branching noise (Q1888824)

Nonnegative solutions of the one-dimensional stochastic partial differential equation \[ \partial_tu=\partial_{xx}u+\theta u-u^2+\sqrt{2u}\,dW \] are studied, where \(W\) is a space-time white noise. It is known that for \(\theta\) sufficiently large the time \(\tau\) of attaining \(0\) by the solution is infinite with positive probability. It is proved that the function \(g\mapsto P_g(\tau<\infty)\) (\(P_g\) being the probability corresponding to the starting point \(g\), where \(g\) is a nonnegative continuous function with compact support) is the Laplace transform of a unique stationary distribution of the equation. A sufficient condition for a measure \(\nu\) to belong to the domain of attraction of this distribution is also given.