Level of noises and long time behavior of the solution for space-time fractional SPDE in bounded domains (Q6066310)

An equation \[ \partial^\beta_t u=-(-\Delta)^\frac{\alpha}{2}u+I^{1-\beta}\left[\lambda\sigma(u)\dot W\right] \tag{1} \] is considered on a regular open bounded set \(B\subseteq\mathbb{R}^d\) with the Dirichlet boundary conditions and with a non-random bounded initial condition \(u_0(x)\). Here \(\beta\in(0,1)\), \(\alpha\in(0,2]\), \(\partial^\beta_t\) denotes the Caputo fractional derivative, \(I^\gamma_t\) denotes the fractional order integral \[ I^\gamma_tf(t)=\frac{1}{\Gamma(\gamma)}\int_0^t(t-\tau)^{\gamma-1}f(\tau)\,d\tau, \] \(\lambda\) is a positive scalar, \(\sigma:\mathbb{R}\to\mathbb{R}\) is a globally Lipschitz function such that \(\ell|x|\le|\sigma(x)|\le L|x|\) for some \(\ell,L>0\) and \(\dot W\) is either white both in space and time, or white in time and colored in space (in the latter case, there are some restrictions on the spatial covariance of the noise). If \(G^{(\beta)}_B\) denotes the space-time fractional kernel of the problem \[ \partial^\beta_t u=-(-\Delta)^\frac{\alpha}{2}u \] with the Dirichlet boundary conditions on \(B\) then a mild solution of (1) is defined as \[ u(t,x)=(G^{(\beta)}_Bu_0)_t(x)+\lambda\int_B\int_0^tG^{(\beta)}_B(t-s,x,y)\sigma(u(s,y))W(ds,dy). \] The authors study the dependence of the long time behavior of second moments (and suprema of second moments over the domain \(B\)) of the mild solution \(u(t,x)\) on the parameters \(\alpha\), \(\beta\), \(\lambda\), \(d\) and on spectral properties of the Dirichlet Laplace operator on \(B\). If the eigenvectors of Dirichlet Laplace operator on \(B\) are uniformly bounded then it is also proved that \[ \lim_{\gamma\to\beta}\sup_{x\in B}\mathbb{E}\,|u^{(\gamma)}(t,x)-u^{(\beta)}(t,x)|^p \] holds for every \(p\ge 2\) in the range \(\beta,\gamma\in(\frac{1}{2},1)\), \(d<\frac{1}{2}\min\,\{\frac{1}{\beta},\frac{1}{\gamma}\}\alpha\) where \(\beta\mapsto u^{(\beta)}\) denotes the dependence of the solution on the fractional parameter \(\beta\).