Intermittency and multifractality: a case study via parabolic stochastic PDEs (Q682245)

``Intermittency'' and ``multifractality'' are two properties which the solutions to stochastic partial differential equations can have. In the present article, the authors show that these two concepts do not coincide. \par For this purpose, they consider the nonlinear stochastic heat equation \[ \dot{u} = \frac{1}{2} u'' + u \xi, \quad u(0) = 1 \] and linear heat equation \[ \dot{Z} = \frac{1}{2} Z'' + \xi, \quad Z(0) = 0. \] It is well known that the solution $u$ to the first equation is intermittent, whereas the solution $Z$ to the second equation is not intermittent. The authors prove that the tall peaks of the solution $u$ to the nonlinear stochastic heat equation are multifractal (which follows from Theorem 1.2), and that the tall peaks of the solution $Z$ to the linear heat equation are multifractal as well (which is a consequence of Theorem 1.3). This demonstrates that the two terms ``intermittency'' and ``multifractality'' are indeed not the same. The mentioned results (Theorems 1.2 and 1.3) are consequences of more general results about multifractal random fields, namely Theorems 4.1 and 4.7, which provide general upper and lower bounds for the Hausdorff dimension of the tall peaks of a stochastic process. \par Moreover, the authors utilize their methods in order to prove that the tall peaks of a Brownian motion are monofractal, whereas the tall peaks of an Ornstein-Uhlenbeck process are multifractal.