Abstract:We consider upper level-sets of the Gaussian free field on $\mathbb Z^d$, for $d\geq 3$, above a given real-valued height parameter $h$. As $h$ varies, this defines a canonical percolation model with strong, algebraically decaying correlations. We prove that three natural critical parameters associated to this model, respectively describing a well-ordered subcritical phase, the emergence of an infinite cluster, and the onset of a local uniqueness regime in the supercritical phase, actually coincide. At the core of our proof lies a new interpolation scheme aimed at integrating out the long-range dependence of the Gaussian free field. Due to the strength of correlations, its successful implementation requires that we work in an effectively critical regime. Our analysis relies extensively on certain novel renormalization techniques that bring into play all relevant scales simultaneously. The approach in this article paves the way to a complete understanding of the off-critical phases for strongly correlated disordered systems.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is about the consistency of several natural critical parameters in the Gaussian Free Field (GFF) level - set percolation model. Specifically, the author studies the percolation properties of the level set $\{ \phi \geq h \}$ of the massless Gaussian free field $\phi$ on $\mathbb{Z}^d$, where $d \geq 3$. This model has strong algebraic decay correlations, making its study very challenging.
The main contribution of the paper is to prove that the three key critical parameters $h^*$, $h^{**}$ and $\bar{h}$ are actually the same. These three parameters describe the following three phases respectively:
1. **Sub - critical phase**: In this phase, the system is in a well - ordered state.
2. **Emergence of infinite clusters**: In this phase, infinitely large connected clusters begin to appear in the system.
3. **Local uniqueness in the super - critical phase**: In this phase, the infinitely large connected clusters in the system become unique.
Specifically, the paper proves that:
\[ \bar{h}(d) = h^*(d) = h^{**}(d) \]
The significance of this result is that it provides a complete framework for understanding the non - critical phases of strongly correlated disordered systems. By introducing new interpolation schemes and renormalization techniques, the author successfully solves the consistency problem of these parameters, which was an open problem in previous research.
### Main Results
- **Theorem 1.1**: For all $d \geq 3$, we have $\bar{h}(d) = h^*(d) = h^{**}(d)$.
- **Corollary 1.2**: For all $d \geq 3$ and $\epsilon > 0$, there exist constants $c = c(d, \epsilon) \in (0, \infty)$ and $\rho = \rho(d) \in (0, \infty)$ such that for all $h \notin (h^* - \epsilon, h^* + \epsilon)$ and $x, y \in \mathbb{Z}^d$, the truncated two - point function $\tau_h(x, y)$ satisfies:
\[
\tau_h(x, y)=P[x \phi \geq h \leftrightarrow y, x \phi \geq h \nleftrightarrow \infty] \leq e^{-c |x - y|^\rho}
\]
### Key Strategies
- **Interpolation and Renormalization**: The core strategy of the paper is to remove the long - range dependence of the model by the interpolation method and to deal with the correlations on different scales by the renormalization technique. Specific steps include:
- Decompose the GFF into a series of independent stationary Gaussian fields $\xi_\ell$.
- Define the truncated field $\phi_L$ and introduce the noisy version $T_\delta \omega$.
- By comparing the critical parameters of the finite - range model and the complete GFF model, gradually prove the consistency of these parameters.
### Conclusion
Through the above methods, the author not only proves the consistency of the critical parameters, but also provides important tools and methods for understanding the geometric properties of strongly correlated disordered systems. This result is of great significance for research in the fields of statistical physics and probability theory.