Percolation (mathematics)

Percolation (mathematics)
Name	Percolation (mathematics)
Caption	Lattice percolation example
Field	Probability theory
Introduced	1950s
Notable	Harry Kesten, Geoffrey Grimmett, Oded Schramm, Stanislav Smirnov

Percolation (mathematics) Percolation studies connectivity and phase transitions in random media using probabilistic, combinatorial, and geometric tools. It links lattice models, stochastic processes, statistical mechanics, and complex analysis through questions about clusters, thresholds, and scaling limits.

Introduction

Percolation emerged in the 1950s and 1960s in works connected to John von Neumann, Stanislaw Ulam, Paul Erdős, Pat Moran, and Solomon Lefschetz, later advanced by Harry Kesten, Geoffrey Grimmett, Oded Schramm, and Stanislav Smirnov. It sits alongside models like the Ising model, Potts model, Random Graphs of Erdős–Rényi, and Self-avoiding walk in the study of statistical mechanics and combinatorics, and interacts with developments by Andrei Kolmogorov, Andrey Markov, Andrey Kolmogorov's school, and researchers at institutions such as Princeton University, Cambridge University, and Institut des Hautes Études Scientifiques. Percolation theory has influenced and been influenced by breakthroughs involving the Renormalization Group, Conformal Field Theory, and methods used by researchers at Université Paris-Sud and Courant Institute.

Definitions and Models

Basic models include bond percolation and site percolation on graphs like the square lattice, triangular lattice, hexagonal lattice, and higher-dimensional cubic lattices studied by groups at Massachusetts Institute of Technology and University of Oxford. In bond percolation each edge of a graph such as Z^d is independently declared open with probability p and closed otherwise; site percolation declares vertices open with probability p. Continuum percolation, studied by collaborators at Bell Labs and Bellcore, models random geometric graphs like the Gilbert disk model, Boolean models, and Poisson blob processes anchored in work by Stuart Rice and Joseph Mecke. Dependent models include invasion percolation, first-passage percolation developed with contributions from Hammersley and Welsh, and bootstrap percolation linked to cellular automata considered by John Conway's community. Percolation is also defined on nonamenable graphs, Cayley graphs studied by Mikhael Gromov, and fractal sets examined in research connected to Benoît Mandelbrot.

Critical Phenomena and Phase Transition

A central object is the critical probability p_c separating almost-sure absence of an infinite cluster from its presence, a concept explored in parallel to the Curie–Weiss model and Kosterlitz–Thouless transition. Results on uniqueness of the infinite cluster and sharpness of phase transition were influenced by techniques from Oded Schramm and Geoffrey Grimmett and connect to universality conjectures considered by Michael Fisher and Kenneth Wilson. Two-dimensional percolation on the triangular lattice attains exact values for p_c proved by methods related to duality and explored by Harry Kesten and Stanislav Smirnov, linking to conformal invariance themes advanced by Alexander Polyakov and Gregory Moore.

Mathematical Results and Theorems

Key theorems include Kesten's result on p_c(Z^2) = 1/2, Russo's formula, the RSW (Russo–Seymour–Welsh) theorem developed with contributions from Lucio Russo and Peter Geoffrey Whittle's academic network, and sharp threshold results by Jeffrey Beckner's contemporaries and later formalized by Ehud Friedgut and Gil Kalai. Scaling relations, critical exponents, and universality statements were rigorously established in special cases by Stanislav Smirnov and Oded Schramm through the introduction of Schramm–Loewner evolution (SLE) which relates to work by Wendelin Werner and Greg Lawler. Results in high dimensions used lace expansion techniques developed by Timothy Hara and Gordon Slade with antecedents in the study of Alexander Grothendieck's categorical methods in topology and algebraic geometry indirectly informing combinatorial techniques. The uniqueness of the infinite cluster on transitive amenable graphs involves input from Geoffrey Grimmett and Itai Benjamini with later refinements by Rodolfo G. Pinsky's collaborators.

Methods and Techniques

Analytic and probabilistic tools include coupling arguments developed in the tradition of Stein's method proponents at University of Chicago, correlation inequalities such as the FKG inequality rooted in work by Michael Fortuin and Chris Kasteleyn communities, differential inequalities like Russo's formula, and martingale methods linked to Joseph Doob's legacy. Geometric and combinatorial techniques use planar duality tied to ideas from William Thurston's circle packing program and conformal mappings related to Riemann and later exploited by Lionel Levine and Yuval Peres. Renormalization and block arguments reflect approaches from Kenneth Wilson's school, while lace expansion and diagrammatic methods owe to collaborations including Timothy Hara and Gordon Slade. Computational and numerical approaches trace to work at Los Alamos National Laboratory and IBM Research applying Monte Carlo methods refined by researchers influenced by Stanislaw Ulam and Nicholas Metropolis.

Applications and Related Topics

Percolation informs porous media studies in contexts associated with Stanford University and Imperial College London research groups, network resilience problems considered by Erdős-related graph theory circles, epidemiological models linked to Centers for Disease Control and Prevention collaborations, and materials science programs at Argonne National Laboratory. It intersects with random planar maps studied by Oded Schramm's collaborators, continuum scaling limits related to Fields Medal work awarded to contributors, and algorithmic aspects developed at MIT and ETH Zurich. Connections exist with electrical network theory rooted in George Green's legacy, with the study of random walks on graphs advanced by Frank Spitzer and Doyle and Snell-influenced authors, and with percolation thresholds relevant to telecommunications infrastructures like those considered by Bell Labs.

Category:Probability theory