Abstract:In this article, we consider the mollified stochastic heat equation with flat initial data and the corresponding Kardar-Parisi-Zhang (KPZ) equation in dimension $d \geq 3$ when the mollification parameter is turned off. In addition, we assume that the strength of the space-time white noise is small enough, such that the solution of the mollified stochastic heat equation has a finite limiting second moment. In this case, with regard to the mollified KPZ equation, we show that the fluctuation of the mollified KPZ equation about its stationary solution converges in the sense of finite dimensional distributions towards a Gaussian free field. On the other hand, about the mollified stochastic heat equation, we show that the fluctuation of the mollified stochastic heat equation about its stationary solution converges in the sense of finite dimensional convergence towards the product of an i.i.d random field and an independent Gaussian free field. Here the i.i.d random field is generated by the limiting partition function of the continuous directed polymer. As an application, we prove the limiting fluctuations of the partition function and the free energy of the continuous directed polymer.
What problem does this paper attempt to address?
The main problem that this paper attempts to solve is to study the spatio - temporal fluctuation behaviors of the Stochastic Heat Equation (SHE) and the Kardar - Parisi - Zhang (KPZ) equation in the entire \(L^2\) region in high - dimensional spaces (dimension \(d\geq3\)). Specifically, the author focuses on the following two aspects:
1. **Fluctuations of SHE and KPZ equations**: When the smoothing parameter is turned off, consider the smoothed stochastic heat equation with flat initial data and its corresponding KPZ equation. Assume that the intensity of the space - time white noise is small enough so that the solution of the smoothed stochastic heat equation has a finite second - order moment. Under this condition, the author proves that:
- For the smoothed KPZ equation, its fluctuations with respect to the stationary solution converge to the Gaussian free field in the sense of finite - dimensional distributions.
- For the smoothed stochastic heat equation, its fluctuations with respect to the stationary solution converge to the product of an independent Gaussian free field and an independent and identically distributed (i.i.d.) random field in the sense of finite - dimensional distributions, where the i.i.d. random field is generated by the limit partition function of the continuous directed polymer.
2. **Asymptotic behavior of continuous directed polymers**: As an application of the above results, the author proves the limit fluctuations of the partition function and free energy of continuous directed polymers.
### Formula Summary
To better understand these problems, the key formulas involved include:
- **Stochastic Heat Equation (SHE)**:
\[
\partial_t U(x,t)=\frac{1}{2}\Delta U(x,t)+\beta U(x,t)\xi(x,t),\quad(x,t)\in\mathbb{R}^d\times\mathbb{R}_+
\]
- **KPZ equation**:
\[
\partial_t H(x,t)=\frac{1}{2}\Delta H(x,t)+\frac{1}{2}|\nabla H(x,t)|^2+\beta\xi(x,t),\quad(x,t)\in\mathbb{R}^d\times\mathbb{R}_+
\]
- **Cole - Hopf transformation**:
\[
H(x,t)=\log U(x,t)
\]
- **Smoothed SHE and KPZ equations**:
\[
\partial_t U_\epsilon(x,t)=\frac{1}{2}\Delta U_\epsilon(x,t)+\beta_\epsilon U_\epsilon(x,t)\xi_\epsilon(x,t),\quad U_\epsilon(0,x)\equiv1
\]
\[
\partial_t H_\epsilon(x,t)=\frac{1}{2}\Delta H_\epsilon(x,t)+\frac{1}{2}|\nabla H_\epsilon(x,t)|^2 - C_\epsilon+\beta_\epsilon\xi_\epsilon(x,t)
\]
- **Fluctuation definitions**:
\[
F^{\text{SHE}}_\epsilon(x,t):=\epsilon^{-\frac{d + 2}{2}}(U^{\text{st}}_\epsilon(x,t)-U_\epsilon(x,t))
\]
\[
F^{\text{KPZ}}_\epsilon(x,t):=\epsilon^{-\frac{d + 2}{2}}(H^{\text{st}}_\epsilon(x,t)-H_\epsilon(x,t))
\]