Abstract:We prove strong rate resp. weak rate ${\mathcal O}(\tau)$ for a structure preserving temporal discretization (with $\tau$ the step size) of the stochastic Allen-Cahn equation with additive resp. multiplicative colored noise in $d=1,2,3$ dimensions. Direct variational arguments exploit the one-sided Lipschitz property of the cubic nonlinearity in the first setting to settle first order strong rate. It is the same property which allows for uniform bounds for the derivatives of the solution of the related Kolmogorov equation, and then leads to weak rate ${\mathcal O}(\tau)$ in the presence of multiplicative noise. Hence, we obtain twice the rate of convergence known for the strong error in the presence of multiplicative noise.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the weak error analysis of the time - discretization scheme for the stochastic Allen - Cahn equation. Specifically, the authors Dominic Breit and Andreas Prohl strive to prove that in 1, 2, and 3 - dimensional spaces, for the stochastic Allen - Cahn equation with additive or multiplicative colored noise, a structure - preserving time - discretization method (where $\tau$ is the step size) can achieve strong and weak convergence rates of $O(\tau)$.
### Background and Motivation
The stochastic Allen - Cahn equation is a phase - field model for describing the dynamics of material interfaces, and its deterministic version has been widely studied. However, when stochastic perturbations are introduced, the solution of the equation becomes more complex. The main objective of this paper is to conduct a detailed weak error analysis of the time - discretization scheme under different types of noise (additive noise and multiplicative noise).
### Main Contributions
1. **Strong Convergence Rate**:
- For additive noise, the authors prove that the time - discretization scheme can achieve a strong convergence rate of $O(\tau)$.
- For multiplicative noise, by using the Kolmogorov equation and high - order discrete stability, the authors also prove that the weak convergence rate is $O(\tau)$.
2. **Technical Details**:
- The authors use a transformation $y(t)=u(t)-\Phi W(t)$ to transform the original equation into a new form, thereby simplifying the strong error analysis.
- For multiplicative noise, the authors use the Kolmogorov equation and high - order discrete stability to derive weak error estimates.
3. **Numerical Experiments**:
- The authors also provide numerical experiment results to verify the correctness of the theoretical analysis.
### Formulas and Techniques
- **Stochastic Allen - Cahn Equation**:
\[
du = (\Delta u-(u^{3}-u))dt+\Phi(u)dW\quad\text{in }Q_T=(0, T)\times T^{d}
\]
where $u(0) = u_0$ on $T^{d}$, $T^{d}=((-\pi,\pi)\vert\{-\pi,\pi\})^{d}$, and $T > 0$, $d = 1, 2, 3$.
- **Energy Functional**:
\[
E(\phi)=\frac{1}{2}\int_{T^{d}}|\nabla\phi|^{2}dx+\int_{T^{d}}F(\phi)dx,\quad\text{where}\quad F(x)=\frac{1}{4}(|x|^{2}-1)^{2}
\]
- **Gradient Flow**:
\[
\partial_{t}u=-D E(u)=\Delta u - f(u)\quad\text{in }L^{2}(T^{d})
\]
where $D E(u)$ represents the Fréchet derivative of $E$ at $u$.
- **Discretization Scheme**:
\[
\frac{u_{m}-u_{m - 1}}{\tau}-\Delta u_{m}+f(u_{m})=\Phi(u_{m - 1})\Delta_{m}W\quad\text{for }m\geq1,\quad u_{0}=u_{0}
\]
where $\Delta_{m}W = W(t_{m})-W(t_{m - 1})$.
### Conclusion
This paper, through a detailed weak error analysis of the time - discretization scheme for the stochastic Allen - Cahn equation, proves that under different types of noise, this scheme can achieve a convergence rate of $O(\tau)$. These results provide an important theoretical basis for understanding and applying stochastic phase - field models.