Inverse problem for semidiscrete stochastic parabolic operators in arbitrary dimensions (2509.03760v1)

Abstract: In this paper, we study two types of inverse problems for space semidiscrete stochastic parabolic equations in arbitrary dimensions. The first problem concerns a semidiscrete inverse source problem, which involves determining the random source term of the white noise in the stochastic parabolic equation using observation data of the solution at the terminal time and the trace of the discrete spatial derivative of the solution on a subdomain at the boundary of the space. The second problem addresses a semidiscrete Cauchy inverse problem, which involves determining the solution of the stochastic parabolic equation in a subdomain of space-time, from the data of the solution and the trace of the discrete spatial derivative on a subdomain of the boundary of space and time. To address these problems, we establish two Carleman estimates for a semidiscrete stochastic parabolic operator in arbitrary dimensions: one for the homogeneous boundary condition and the other for the nonhomogeneous boundary condition. Applying these Carleman estimates, we obtain Lipschitz and H\"older stability for the first and second inverse problems, respectively.