Lower tail of the KPZ equation (Q2178469)

The paper deals with the stochastic heat equation with multiplicative space-time white noise, which can be formally leaded to the Kardar-Parisi-Zhang (KPZ) equation. Considering narrow wedge initial data, tight bounds are established on the lower tail probability that the solution is close to zero. In particular, the authors find that the upper bound for the lower tail probability of the centered and scaled KPZ solution can be written as the sum of three terms, the first one has a superexponential decay with exponent 5/2 and dominates for tail depth greater than \(T^{3/2}\), with \(T\) the time, the last one has exponent 3 (that is the tail behavior of the Gaussian unitary ensemble Tracy-Widom distribution is recovered) and dominates for tail depth less than \(T^{3/2}\) and \(T\) going to infinity. A lower bound is also provided.