Abstract:A stochastic SIR epidemic model taking into account the heterogeneity of the spatial environment is constructed. The deterministic model is given by a partial differential equation and the stochastic one by a space-time jump Markov process. The consistency of the two models is given by a law of large numbers. In this paper, we study the deviation of the spatial stochastic model from the deterministic model by a functional central limit theorem. The limit is a distribution-valued Ornstein-Uhlenbeck Gaussian process, which is the mild solution of a stochastic partial differential equation.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to study the deviation between the stochastic SIR epidemic model and the deterministic model on the spatially discrete grid. Specifically, the authors focus on how the stochastic model deviates from its corresponding deterministic model as the grid step size approaches zero, and describe this deviation through the Functional Central Limit Theorem (FCLT).
### Main problems
1. **Model construction**: The paper first constructs a stochastic SIR epidemic model considering spatial heterogeneity, which is based on the space - time jump Markov process.
2. **Consistency verification**: Through the Law of Large Numbers, the consistency between the stochastic model and the deterministic model is verified.
3. **Deviation analysis**: The deviation between the stochastic model and the deterministic model when the grid step size approaches zero is studied, and the asymptotic distribution of the deviation is given through the Functional Central Limit Theorem.
### Mathematical description
- **Deterministic model**: Described by partial differential equations (PDEs), expressed as:
\[
\begin{cases}
\frac{\partial s}{\partial t}(t,x) = \mu_S \Delta s(t,x) - \beta(x) \frac{s(t,x)i(t,x)}{s(t,x) + i(t,x) + r(t,x)} \\
\frac{\partial i}{\partial t}(t,x) = \mu_I \Delta i(t,x) + \beta(x) \frac{s(t,x)i(t,x)}{s(t,x) + i(t,x) + r(t,x)} - \alpha(x) i(t,x) \\
\frac{\partial r}{\partial t}(t,x) = \mu_R \Delta r(t,x) + \alpha(x) i(t,x)
\end{cases}
\]
- **Stochastic model**: Described by the space - time jump Markov process, expressed as:
\[
Z_N^\varepsilon(t) = Z_N^\varepsilon(0) + \int_0^t b_\varepsilon(r, Z_N^\varepsilon(r)) dr + \frac{1}{\sqrt{N}} \sum_{j = 1}^K h_j P_j \left( N \int_0^t \beta_j(Z_N^\varepsilon(r)) dr \right)
\]
### Functional Central Limit Theorem
As the grid step size \(\varepsilon\) approaches zero, the deviation of the stochastic model can be described by a distribution - valued Ornstein - Uhlenbeck Gaussian process, which is the mild solution of the following stochastic partial differential equation (SPDE):
\[
\Psi_\varepsilon(t) = \int_0^t \nabla_z b_\varepsilon(r, Z_\varepsilon(r)) \Psi_\varepsilon(r) dr + \sum_{j = 1}^K \int_0^t \sqrt{\beta_j(r, Z_\varepsilon(r))} dB_j(r)
\]
### Conclusion
The main contribution of the paper is that through strict mathematical derivations, it is proved that under certain conditions, as the grid step size approaches zero, the deviation of the stochastic SIR model can be described by a Gaussian process, which provides a theoretical basis for understanding the impact of stochastic fluctuations on epidemic spread.