Abstract:We obtain the precise Fourier dimension of the Gaussian multiplicative chaos on the unit interval. Our main result confirms a conjecture of Garban-Vargas.
What problem does this paper attempt to address?
### The problems the paper attempts to solve
This paper aims to solve the Fourier dimension problem of Gaussian Multiplicative Chaos (GMC). Specifically, the authors strive to accurately calculate the Fourier dimension of GMC on the unit interval and verify the Garban - Vargas conjecture.
#### Background Introduction
Gaussian Multiplicative Chaos (GMC), a random measure theory introduced by Kahane in the 1980s. The GMC measure can be regarded as the exponential form of the log - correlated Gaussian field. The GMC measure plays an important role in fields such as mathematical physics, Liouville quantum gravity, and multifractal analysis. Due to its fundamental significance, the harmonic analysis properties of the GMC measure (such as whether it can be classified as a Salem measure) have attracted wide attention.
In recent years, the harmonic analysis, Fourier decay, and Fourier dimension of GMC have received special attention. For example, Falconer and Jin [FJ19] provided a non - trivial lower bound for two - dimensional GMC under the small parameter \(\gamma < \frac{1}{\sqrt{3}}\). Garban and Vargas [GV24] studied the Rajchman property of the standard GMC measure \(M_\gamma\) on the unit circle and gave a lower bound for the Fourier dimension for the small parameter \(\gamma < \frac{1}{\sqrt{2}}\):
\[ \frac{1}{2}-\gamma^{2}\leq\dim_F(M_\gamma)\leq1 - \gamma^{2}<\dim_H(M_\gamma)=1-\frac{\gamma^{2}}{2}. \]
Based on the results of rescaled fluctuations, Garban and Vargas conjectured that for the small parameter \(\gamma < \frac{1}{\sqrt{3}}\), the Fourier dimension of the GMC measure \(M_\gamma\) is \(1-\gamma^{2}\). In addition, they also raised a question: whether the Fourier dimension of the GMC measure is given by its correlation dimension in all sub - critical parameter ranges.
#### Main Results
This paper solves the Garban - Vargas conjecture and proves that for \(\gamma\in(0,\sqrt{2})\) in the sub - critical parameter range, the Fourier dimension of the GMC measure \(\mu_{\gamma,\text{GMC}}\) is \(D_\gamma\), where:
\[ D_\gamma:=\begin{cases}
1-\gamma^{2}&\text{if }0 < \gamma < \frac{\sqrt{2}}{2},\\
(\sqrt{2}-\gamma)^{2}&\text{if }\frac{\sqrt{2}}{2}\leq\gamma < \sqrt{2}.
\end{cases} \]
The main results are as follows:
**Theorem 1.1**: For each \(\gamma\in(0,\sqrt{2})\), almost surely \(\dim_F(\mu_{\gamma,\text{GMC}})=D_\gamma\).
#### Proof Sketch
The main part of the proof is to establish the almost sure lower bound \(\dim_F(\mu_{\gamma,\text{GMC}})\geq D_\gamma\). The upper bound is already known from Lacoin - Rhodes - Vargas and Garban - Vargas [LRV15] and [GV24].
To establish the lower bound \(\dim_F(\mu_{\gamma,\text{GMC}})\geq D_\gamma\), the authors use the vector - valued martingale method, in particular, the Bacry - Muzy white noise decomposition method to construct GMC. By defining a series of independent random processes and studying their limit behavior, the authors finally prove the required inequality.