The non-Archimedean stochastic heat equation driven by Gaussian noise (Q498957)

This paper treats the non-Archimedean stochastic heat equation driven by Gaussian noise of the type \[ \begin{aligned} & \frac{\partial u}{\partial t}(t,x) + A( \partial, \beta) u(t,x) = \sigma ( u(t,x)) \dot{W}(t,x) + b( u(t,x)), \\ & u(0,x) = u_0(x), \qquad t \geq 0, \quad x \in {\mathbb Q}_p^N, \end{aligned}\tag{1} \] where the coefficients \(\sigma\) and \(b\) are real-valued functions and \(\dot{W}(t,x)\) is the formal notation for the Gaussian random perturbation, \({\mathbb Q}_p\) is the field of \(p\)-adic numbers with a prime number \(p\), and \({\mathbb Q}_p^N\) denotes a complete ultrametric space with the \(p\)-adic norm \(\| \cdot \|_p\). For a fixed \(\beta> 0\), \[ A( \partial, \beta) \phi(x) = {\mathcal F}_{ \xi \to x}^{-1} ( | a (\xi) |_p^{\beta} {\mathcal F}_{ x \to \xi} \phi), \qquad \phi \in {\mathcal D}, \tag{2} \] is an elliptic pseudodifferential operator of degree \(d\) with symbol \(| a |_p^{\beta}\), where \(a(\xi)\) is an elliptic polynomial of degree \(d\) and satisfies \[ C_0 \| \xi \|_p^d \leqslant | a(\xi) |_p \leqslant C_1 \| \xi \|_p^d, \qquad \text{ for all }\xi \in {\mathbb Q}_p^N \tag{3} \] for some positive constants \(C_0, C_1> 0\) [\textit{W. A. Zúñiga-Galindo}, Potential Anal. 28, No. 2, 185--200 (2008; Zbl 1134.35005)]. The goal of this paper is to prove the existence and uniqueness of a mild random field solution \(u(t,x)\) of the associated stochastic integral equation \[ \begin{aligned} & u(t,x) = ( \Gamma(t) * u_0) (x) + \int_0^t \int_{ {\mathbb Q}_p^N } \Gamma(t-s, x-y) \sigma( u(s,y)) W(ds, d^N y) \\ & + \int_0^t ds \int_{ {\mathbb Q}_p^N} \Gamma(s,y) b( u(t-s, x-y))\, d^N y, \quad \text{a.s.} \quad \text{for all} \quad (t,x) \in [0, T] \times {\mathbb Q}_p^N. \end{aligned}\tag{4} \] The heat kernel or fundamental solution \(\Gamma(t,x)\) is given by \[ \Gamma(t,x) := {\mathcal F}_{ \xi \to x}^{-1} ( e^{ -t | a(\xi) |_p^{\beta} } ) \qquad \text{for} \quad t>0, \quad x \in {\mathbb Q}_p^N.\tag{5} \] Assume that \[ \int_0^T dt \int_{ {\mathbb Q}_p^N } | {\mathcal F} \Gamma(t) (\xi) |^2\, d \mu(\xi) < \infty \quad \text{and} \quad \sup_{ 0 \leqslant t \leqslant T} \Gamma(t, {\mathbb Q}_p^N ) < \infty \tag{6} \] with the associated measure \(\Gamma(t, d^N x) := \Gamma(t,x) d^N x\), where \(d^N x\) is the Haar measure on \({\mathbb Q}_p^N\) and \(\mu\) is a regular Borel measure on \({\mathbb Q}_p^N\). Assume also that \[ \lim_{ h \to o+} \int_0^T \int_{ {\mathbb Q}_p^N } \sup_{ | r-t | < h} | {\mathcal F} \Gamma(r)(\xi) - {\mathcal F} \Gamma(t)(\xi) |^2 d \mu(\xi)\, dt = 0. \tag{7} \] Here is the main result: if \(b, \sigma\) are Lipschitz continuous and \(u_0\) is measurable and bounded, and if \[ \int_{ {\mathbb Q}_p^N } \| \xi \|_p^{ d \beta}\, d \mu(\xi)< \infty, \tag{8} \] then there exists a unique mild random field solution \(\{ u(t,x) \}\), \((t,x)\) \(\in\) \([0, T] \times {\mathbb Q}_p^N\), of the equation (4), and \(u\) is \(L^2( \Omega)\)-continuous and satisfies \[ \sup_{ (t,x) \in [0, T] \times {\mathbb Q}_p^N } E ( | u(t,x) |^2 )< \infty. \tag{9} \] The stochastic integral term is defined as a stochastic integration with respect to a spatially homogeneous Gaussian noise in the sense of, e.g., [\textit{R. C. Dalang}, Electron. J. Probab. 4, Paper No. 6, 29 p. (1999; Zbl 0922.60056)]. For other related works, see, e.g., [\textit{A. N. Kochubei}, Pseudo-differential equations and stochastics over non-Archimedean fields. New York, NY: Marcel Dekker (2001; Zbl 0984.11063)] and [\textit{J. J. Rodríguez-Vega} and \textit{W. A. Zúñiga-Galindo}, Pac. J. Math. 237, No. 2, 327--347 (2008; Zbl 1232.35201)].