Poisson-Dirichlet statistics for the extremes of a log-correlated Gaussian field (Q2511555)

In the present paper, the statistics of the extremes of a Gaussian field whose correlations decay logarithmically with the distance is studied. At low temperature, it is shown that the normalized covariance of two points sampled from the Gibbs measure is either 0 or 1. Using this, it is proved that the joint distribution of the Gibbs weights converges in a suitable sense to that of a Poisson-Dirichlet variable. The method of proof is robust and applicable to a large class of nonhierarchical log-correlated fields.