Turbulence statistics at two points — the velocity difference between locations separated by distance r — are well characterized by structure functions. But turbulence is a multi-scale, multi-point phenomenon. Three-point statistics capture how energy transfers connect scales. Four-point and higher statistics characterize the geometry of turbulent structures. These multipoint statistics have remained largely unexplored theoretically because they are computationally prohibitive to measure and analytically intractable to derive.
A closure of the N-point velocity increment Hopf equation provides the analytical framework (arXiv:2603.11595). The Hopf equation is the master equation for turbulence — it governs the probability density functional of the entire velocity field. In velocity-increment variables, the equation describes how the joint statistics of velocity differences at N points evolve. The closure, generalizing a single-point proposal by Sreenivasan and Yakhot, makes the N-point equation tractable.
The first application: the transition between the known two-point structure function and the three-point fusion rules. The analytical solution takes the form of a Batchelor interpolation — a smooth connection between the two limiting behaviors. The result agrees with preliminary direct numerical simulation data.
The significance is structural. The closure is not a model — it is a first-principles-based approximation that preserves the Hopf equation's structure. Because the N-point equation is closed, it can be solved numerically for arbitrary N, and analytical predictions can be extracted for specific multipoint quantities using the same methods that produced the three-point result.
Turbulence has a hierarchy of statistics. The hierarchy now has a closed equation at every level.