Abstract:This paper proposes and analyzes a new operator splitting method for stochastic Maxwell equations driven by additive noise, which not only decomposes the original multi-dimensional system into some local one-dimensional subsystems, but also separates the deterministic and stochastic parts. This method is numerically efficient, and preserves the symplecticity, the multi-symplecticity as well as the growth rate of the averaged energy. A detailed $H^2$-regularity analysis of stochastic Maxwell equations is obtained, which is a crucial prerequisite of the error analysis. Under the regularity assumptions of the initial data and the noise, the convergence order one in mean square sense of the operator splitting method is established.
What problem does this paper attempt to address?
The main problem that this paper attempts to solve is to propose and analyze a new operator splitting method for solving the stochastic Maxwell equations with additive noise. Specifically, the paper aims to:
1. **Decompose multi - dimensional systems**: Decompose the original multi - dimensional stochastic Maxwell equations into several local one - dimensional subsystems, thereby significantly reducing the computational complexity and improving the computational efficiency.
2. **Separate the deterministic and stochastic parts**: Separate the deterministic and stochastic parts by the splitting technique, making the solution of each subsystem simpler.
3. **Preserve the geometric structure**: Ensure that the splitting method can preserve the symplectic structure and multi - symplectic conservation law of the original equations, as well as the growth rate of the average energy.
4. **Provide convergence analysis**: Establish the convergence order of the splitting method in the mean - square sense and give a detailed \(H^2\) - regularity analysis as a prerequisite for error analysis.
### Specific problem description
The stochastic Maxwell equations studied in the paper are as follows:
\[
\begin{aligned}
dE(t) &= \text{curl} H(t) dt+\lambda_1 dW(t), \quad(t, x)\in(0, T]\times D, \\
dH(t) &= -\text{curl} E(t) dt+\lambda_2 dW(t), \quad(t, x)\in(0, T]\times D, \\
E(0, x) &= E_0(x), \quad H(0, x)=H_0(x), \quad x\in D,
\end{aligned}
\]
where \(D=(a_1^-, a_1^+)\times(a_2^-, a_2^+)\times(a_3^-, a_3^+)\subseteq\mathbb{R}^3\) is a cubic region, \(E\) and \(H\) represent the electric field and the magnetic field respectively, \(\lambda_i\in\mathbb{R}^3\) describes the amplitude of the noise, \(\{W(t)\}_{t\in[0, T]}\) is a standard \(Q\)-Wiener process with respect to a filtered probability space \((\Omega, \mathcal{F}, \{\mathcal{F}_t\}_{0\leq t\leq T}, P)\), and \(Q\) is a symmetric, positive - definite and trace - finite operator on the Hilbert space \(U\).
### Main contributions
- Proposed a new operator splitting method, which not only decomposes the multi - dimensional system into several local one - dimensional subsystems, but also separates the deterministic and stochastic parts.
- Proved that the convergence order of this splitting method in the mean - square sense is first - order.
- Analyzed in detail the \(H^2\) - regularity of the stochastic Maxwell equations, which is an important prerequisite for error analysis.
Through these works, the paper provides an efficient numerical method for large - scale and long - time simulation of stochastic Maxwell equations.