Technical Summary: Invisible Defect Cascades for Navier–Stokes Regularity

1. Problem Statement

The paper addresses the local regularity problem for the three-dimensional incompressible Navier–Stokes equations. Specifically, it investigates the scenario where a potential singular point $z_0$ fails to enter the Caffarelli–Kohn–Nirenberg (CKN) smallness regime at any sufficiently small dyadic scale. In classical theory, if a scale-invariant quantity $\Psi(z_0, r) = C(z_0, r) + D(z_0, r)$ falls below a universal threshold $\varepsilon_{CKN}$, the solution is regular. The paper assumes the contrary: that for a non-CKN branch, $\Psi(r_n) > \varepsilon_{CKN}$ for all dyadic scales $r_n = 2^{-n}$.

The central question is not to prove regularity directly, but to structurally characterize the nature of the obstruction that prevents regularity. The paper posits that if a singularity exists, it cannot be a generic concentration of energy or pressure; rather, it must take a highly specific form: an NS-realizable, cleaned, scale-critical, combined-invisible defect cascade.

2. Methodology and Framework

The methodology relies on a multi-stage reduction process that transforms the local regularity problem into a compatibility and observability problem. The framework proceeds through the following steps:

2.1 Dyadic Coarse-Grained Defect Packages

The authors associate a "defect package" to every dyadic scale near a potential singularity. This package is derived from the Navier–Stokes equations via rescaling and coarse-graining (mollification).

• Rescaling: The solution $(u, p)$ is rescaled to unit scale: $u^{(n)}(x,t) = r_n u(r_n x, r_n^2 t)$.
• Coarse-Graining: A mollifier $S_\ell$ is applied to generate resolved fields $U_{n,\ell}$ and $P_{n,\ell}$.
• Defect Definition: The difference between the nonlinear flux of the resolved field and the resolved nonlinear flux defines the Reynolds defect $R_{n,\ell} = S_\ell(u^{(n)} \otimes u^{(n)}) - U_{n,\ell} \otimes U_{n,\ell}$.
• Pressure Decomposition: The pressure is split into an active part ($P^{act}$), generated by the local quadratic source, and a harmonic part ($P^{har}$), which is spatially harmonic in the observation window. Crucially, the harmonic part is retained in the package and not quotiented out as a gauge, as it contributes to momentum and energy.
• Energy Flux: The energy flux $\Pi_{n,\ell} = -R_{n,\ell} : \nabla U_{n,\ell}$ is defined to measure interscale energy transfer.

2.2 Combined Observability Maps

The paper introduces a hierarchy of observation channels to test whether a defect is "visible." A defect is considered invisible only if it passes all tests simultaneously. The channels are:

1. Active Pressure: Tests the non-harmonic pressure generated by the active source.
2. Flux: Tests the interscale energy transfer ($\Pi$).
3. Energy: Tests the local energy balance and positive Reynolds covariance ($R \ge 0$).
4. Adjoint Trace: Tests whether the defect can be detected by selected-time perturbations via the adjoint linearized system.

The Combined Observability Map $O_{comb}$ aggregates these channels. A "true defect phantom" is a non-zero direction in the kernel of this map that is also NS-realizable (i.e., it arises from an actual suitable weak solution, not just a formal linearization).

2.3 Moving-Window Growth Control

The framework moves from fixed windows to a sequence of moving windows $W_n$ as the scale $n \to \infty$. The regularity argument depends on whether the observability constants $M_n$ (the inverse of the smallest singular value of the observation map) grow too rapidly. If the growth is controlled (e.g., polynomial or single-exponential) and the sum of inverse constants diverges, a CKN scale must eventually appear unless an invisible cascade exists.

2.4 Critical Recurrence Diagnostics

To explain how a singularity could persist, the paper proposes a "critical recurrence" program. It suggests that a persistent bad orbit must be sustained by one of three mechanisms:

• Vortex Stretching: Positive resolved stretching production.
• Active Pressure Work: Focusing of active pressure.
• Interscale Flux: Sign-coherent or oscillatory energy flux.
  The paper formulates diagnostics for these mechanisms to determine if they can sustain a recurrence without depleting a finite local budget.

3. Key Contributions

1. Structural Reduction Theorem: The paper provides a precise reduction: if a point is singular, then either the moving-window observability is non-effective (constants grow too fast), or there exists an NS-realizable combined-invisible defect cascade. This shifts the problem from "proving regularity" to "ruling out a specific type of highly organized obstruction."
2. NS-Realizability Filter: A novel contribution is the distinction between formal finite-dimensional kernels and genuine Navier–Stokes obstructions. A formal kernel vector is only dangerous if it lies in the closure of the range of actual NS-derived defect packages. This prevents "model phantoms" from being mistaken for genuine singularities.
3. Combined Observability Mechanism: The paper integrates pressure, flux, energy, and trace observations into a single hierarchy. It demonstrates that a defect invisible to one channel is forced into a smaller residual class tested by the next, culminating in a trace-cost analysis.
4. Explicit Failure Objects: The paper defines the precise mathematical objects that would constitute a counterexample: a sequence of windows, NS-realizable defects, and dual directions that are invisible to all channels but pair non-trivially.
5. Four Theorem Targets: The work reduces the path to a regularity proof to four concrete analytic tasks:
   • Dyadic defect extraction.
   • Observable depletion (visible defects must consume a finite budget).
   • Finite-window NS-realizability exclusion.
   • Moving-window growth control.

4. Results and Status

The paper does not prove the global Navier–Stokes regularity conjecture, nor does it construct a singular solution. Its results are structural and conditional:

• Model Verifications: The paper proves that in "clean" regimes (e.g., periodic finite windows or perturbatively localized windows), the combined observability holds, and the only obstructions are explicit finite-dimensional kernels or projected trace-obstruction spaces.
• Conditional Regularity: Theorem 15.5 states that if the four theorem targets are met in a controlled window class, then a CKN scale must occur, implying local regularity.
• The Contrapositive: If regularity fails, the failure must manifest as a "residual left-singular invisible cascade"—a sequence of NS-realizable defects that are invisible to pressure, flux, energy, and trace, and which align with nearly invisible dual directions.

5. Significance and Claims

The paper claims significance as a structural reduction rather than a solution.

• Refinement of the Obstruction: It compresses the vague notion of "compactness failure" into a highly constrained object: a scale-critical, cleaned, NS-realizable cascade that is invisible to a combined set of physical observables.
• Clarification of the Path Forward: By isolating the four theorem targets, the paper clarifies exactly what analytic work is required to prove regularity. It suggests that if the route fails, the failure must be due to a specific, highly organized mechanism (e.g., a specific alignment of vortex stretching and pressure work that evades all budgets), rather than a generic lack of compactness.
• Honest Status: The authors explicitly state that the paper does not announce the exclusion of singularities. Instead, it transforms the question "Can a Navier–Stokes singularity occur?" into the more structured question "Can there exist an NS-realizable invisible moving-window defect cascade?"

In summary, the paper provides a rigorous framework to analyze potential singularities by decomposing them into defect packages and testing them against a hierarchy of physical observables. It concludes that any genuine singularity must be a "cleaned" object that evades detection by pressure, flux, energy, and trace simultaneously while satisfying the exact Navier–Stokes identities, thereby defining the precise nature of the remaining obstruction to regularity.