Quenched Strong Leray–Hopf Solutions

• Quenched strong Leray–Hopf solutions are enhanced weak solutions of the Navier–Stokes equations that, under deterministic or pathwise stochastic settings, are expected to exhibit improved integrability, uniqueness, or energy equality.
• In the deterministic framework, generic Leray–Hopf solutions typically fail to achieve strong regularity unless they satisfy critical scaling conditions, with global regularity ensuring both uniqueness and precise energy behavior.
• In stochastic models, solutions constructed on a fixed Wiener process demonstrate pathwise strong behavior while often exhibiting non-uniqueness, thereby emphasizing the decisive role of energy inequality as a selection principle.

“Quenched strong Leray–Hopf solutions” does not denote a single universal definition in the works considered here. Rather, it points to two closely related viewpoints. In deterministic three-dimensional Navier–Stokes theory, “quenched” refers to Baire generic initial data in $L^2_\sigma$, and the question is whether a typical Leray–Hopf solution enjoys stronger integrability, uniqueness, or energy equality properties. In stochastic fluid models, “quenched” refers to the pathwise or probabilistically strong setting: the stochastic basis and the Wiener process are fixed, and one studies Leray–Hopf solutions adapted to the given filtration. In both settings, the central issue is whether an energy-class weak solution becomes “strong” in some additional sense, such as belonging to critical function spaces, being unique, satisfying energy equality, being right-continuous in the strong $L^2$ topology, or becoming strong for all positive times (Lindberg, 2024, Lange et al., 2024).

1. Deterministic and stochastic meanings of “quenched” and “strong”

For the three-dimensional incompressible Navier–Stokes equations on $\mathbb{R}^3$,

$$\partial_t u^\nu + (u^\nu \cdot \nabla) u^\nu - \nu \Delta u^\nu + \nabla p^\nu = 0,\qquad \nabla\cdot u^\nu=0,$$

Leray–Hopf solutions are weak solutions satisfying the energy inequality

$$\frac{1}{2}\|u^\nu(t)\|_{L^2}^2+\nu\int_0^t\|\nabla u^\nu(s)\|_{L^2}^2\,ds\le \frac{1}{2}\|u_0\|_{L^2}^2.$$

In the Baire-category framework, a property holds for a Baire generic datum if it holds on a residual set in $L^2_\sigma$, and “quenched” refers to this generic behavior of initial data rather than to probability (Lindberg, 2024).

In stochastic models, the phrase shifts meaning. A solution is probabilistically strong if it is constructed on a fixed stochastic basis with a given Wiener process and is adapted to the given filtration. Several recent works formulate Leray–Hopf solutions precisely as analytically weak, probabilistically strong solutions. In that language, “quenched” is the pathwise viewpoint: the randomness is fixed and the solution is studied for that realization (Berkemeier, 2024, Lange et al., 2024).

The adjective “strong” also has more than one use. In classical PDE terminology it may refer to higher regularity, such as $\partial_t u,\nabla^2u\in L^2$. In other contexts it refers to additional time regularity or critical-space control. The deterministic literature considered here uses “strong Leray–Hopf solution” for Leray–Hopf solutions lying in spaces such as $L^r(0,T;L^s)$ with Ladyzhenskaya–Prodi–Serrin or related scaling, or in Onsager-critical Besov classes. The stochastic literature often uses “probabilistically strong” while the solutions remain analytically weak (Lindberg, 2024, Berkemeier, 2024).

2. Baire-generic Leray–Hopf solutions in 3D Navier–Stokes

The generic theory developed for $u_0\in L^2_\sigma$ studies whether Leray–Hopf solutions belong to Bochner spaces $L^r(0,\infty;X)$, where $L^2$0 is an $L^2$1-homogeneous Banach space. The key scaling quantity is

$L^2$2

If

$L^2$3

then for a Baire generic $L^2$4, no weak solution $L^2$5 belongs to $L^2$6. In the special case $L^2$7, if

$L^2$8

then for a generic $L^2$9 datum there is no weak solution in $\mathbb{R}^3$0 (Lindberg, 2024).

This directly affects the idea of a quenched strong Leray–Hopf solution in the deterministic setting. The results rule out, for a Baire generic $\mathbb{R}^3$1 datum, $\mathbb{R}^3$2 integrability and various other known sufficient conditions for the energy equality. In particular, $\mathbb{R}^3$3 integrability is generically false, and many Shinbrot-type conditions are also generically violated. The paper explicitly interprets this as saying that, generically, Leray–Hopf solutions do not lie in the standard stronger integrability classes that are known to imply smoothness, uniqueness, or energy equality (Lindberg, 2024).

A central consequence is that, without extra hypotheses, the generic Leray–Hopf solution remains only in the basic Leray–Hopf class $\mathbb{R}^3$4. The deterministic “quenched” picture is therefore largely negative: outside very specific critical regimes, strong Leray–Hopf behavior is not typical.

3. Critical scaling, rigidity, and critical Besov regularity

The same deterministic analysis identifies a rigid borderline. In the regime

$\mathbb{R}^3$5

global solvability in $\mathbb{R}^3$6 is equivalent to the a priori estimate

$\mathbb{R}^3$7

Furthermore, one can only have

$\mathbb{R}^3$8

if

$\mathbb{R}^3$9

namely on the Euler scaling line (Lindberg, 2024).

This makes the critical line the only plausible location for any viscosity-uniform quenched strong theory. The paper’s Besov application sharpens this point. For suitable $$\partial_t u^\nu + (u^\nu \cdot \nabla) u^\nu - \nu \Delta u^\nu + \nabla p^\nu = 0,\qquad \nabla\cdot u^\nu=0,$$0-homogeneous Banach spaces $$\partial_t u^\nu + (u^\nu \cdot \nabla) u^\nu - \nu \Delta u^\nu + \nabla p^\nu = 0,\qquad \nabla\cdot u^\nu=0,$$1, each $$\partial_t u^\nu + (u^\nu \cdot \nabla) u^\nu - \nu \Delta u^\nu + \nabla p^\nu = 0,\qquad \nabla\cdot u^\nu=0,$$2 has a Leray–Hopf solution

$$\partial_t u^\nu + (u^\nu \cdot \nabla) u^\nu - \nu \Delta u^\nu + \nabla p^\nu = 0,\qquad \nabla\cdot u^\nu=0,$$3

if and only if a uniform-in-viscosity bound

$$\partial_t u^\nu + (u^\nu \cdot \nabla) u^\nu - \nu \Delta u^\nu + \nabla p^\nu = 0,\qquad \nabla\cdot u^\nu=0,$$4

holds. Moreover, for every such $$\partial_t u^\nu + (u^\nu \cdot \nabla) u^\nu - \nu \Delta u^\nu + \nabla p^\nu = 0,\qquad \nabla\cdot u^\nu=0,$$5, there exists a sequence $$\partial_t u^\nu + (u^\nu \cdot \nabla) u^\nu - \nu \Delta u^\nu + \nabla p^\nu = 0,\qquad \nabla\cdot u^\nu=0,$$6 such that $$\partial_t u^\nu + (u^\nu \cdot \nabla) u^\nu - \nu \Delta u^\nu + \nabla p^\nu = 0,\qquad \nabla\cdot u^\nu=0,$$7 converges weakly$$\partial_t u^\nu + (u^\nu \cdot \nabla) u^\nu - \nu \Delta u^\nu + \nabla p^\nu = 0,\qquad \nabla\cdot u^\nu=0,$$8 in $$\partial_t u^\nu + (u^\nu \cdot \nabla) u^\nu - \nu \Delta u^\nu + \nabla p^\nu = 0,\qquad \nabla\cdot u^\nu=0,$$9 to an Euler weak solution with the same $$\frac{1}{2}\|u^\nu(t)\|_{L^2}^2+\nu\int_0^t\|\nabla u^\nu(s)\|_{L^2}^2\,ds\le \frac{1}{2}\|u_0\|_{L^2}^2.$$0 bound (Lindberg, 2024).

The structural message is that there is no intermediate “generic” strong regularity. Either one has a critical, viscosity-uniform a priori estimate of the correct scaling type, or the set of data producing critical-space strong Leray–Hopf solutions is meagre. This suggests that the deterministic notion of a quenched strong Leray–Hopf solution is intrinsically tied to critical scaling.

4. Conditional quenched strength: uniqueness, energy equality, and inviscid limits

The deterministic generic picture changes drastically under global regularity assumptions. If the Navier–Stokes global regularity statement holds for smooth Schwartz data, then for a Baire generic $$\frac{1}{2}\|u^\nu(t)\|_{L^2}^2+\nu\int_0^t\|\nabla u^\nu(s)\|_{L^2}^2\,ds\le \frac{1}{2}\|u_0\|_{L^2}^2.$$1, the Leray–Hopf solution is unique and satisfies the energy equality for almost every $$\frac{1}{2}\|u^\nu(t)\|_{L^2}^2+\nu\int_0^t\|\nabla u^\nu(s)\|_{L^2}^2\,ds\le \frac{1}{2}\|u_0\|_{L^2}^2.$$2. The paper interprets this as a “quenched-strong” scenario: under global regularity, the generic Leray–Hopf solution is unique, smooth for $$\frac{1}{2}\|u^\nu(t)\|_{L^2}^2+\nu\int_0^t\|\nabla u^\nu(s)\|_{L^2}^2\,ds\le \frac{1}{2}\|u_0\|_{L^2}^2.$$3, and obeys energy equality (Lindberg, 2024).

A parallel statement is obtained for Euler. If for some $$\frac{1}{2}\|u^\nu(t)\|_{L^2}^2+\nu\int_0^t\|\nabla u^\nu(s)\|_{L^2}^2\,ds\le \frac{1}{2}\|u_0\|_{L^2}^2.$$4 the three-dimensional Euler equations are globally well-posed in $$\frac{1}{2}\|u^\nu(t)\|_{L^2}^2+\nu\int_0^t\|\nabla u^\nu(s)\|_{L^2}^2\,ds\le \frac{1}{2}\|u_0\|_{L^2}^2.$$5, then for a Baire generic $$\frac{1}{2}\|u^\nu(t)\|_{L^2}^2+\nu\int_0^t\|\nabla u^\nu(s)\|_{L^2}^2\,ds\le \frac{1}{2}\|u_0\|_{L^2}^2.$$6, anomalous energy dissipation fails in the inviscid limit. More precisely, for generic $$\frac{1}{2}\|u^\nu(t)\|_{L^2}^2+\nu\int_0^t\|\nabla u^\nu(s)\|_{L^2}^2\,ds\le \frac{1}{2}\|u_0\|_{L^2}^2.$$7 there exist sequences $$\frac{1}{2}\|u^\nu(t)\|_{L^2}^2+\nu\int_0^t\|\nabla u^\nu(s)\|_{L^2}^2\,ds\le \frac{1}{2}\|u_0\|_{L^2}^2.$$8 and Leray–Hopf solutions $$\frac{1}{2}\|u^\nu(t)\|_{L^2}^2+\nu\int_0^t\|\nabla u^\nu(s)\|_{L^2}^2\,ds\le \frac{1}{2}\|u_0\|_{L^2}^2.$$9 such that

$L^2_\sigma$0

for all $L^2_\sigma$1, and the limiting Euler solution satisfies

$L^2_\sigma$2

Both the generic energy-equality/uniqueness statement and the no-anomalous-dissipation statement also hold on $L^2_\sigma$3 (Lindberg, 2024).

These results are conditional, but they delineate the positive side of the theory. Without global regularity, generic strong behavior fails in the known supercritical classes. With global regularity, generic Leray–Hopf solutions become strong in the most classical sense available in the energy framework.

5. Pathwise and probabilistically strong Leray–Hopf solutions in stochastic models

In stochastic fluid equations, the term “quenched strong Leray–Hopf solution” is naturally associated with probabilistically strong Leray–Hopf solutions, namely solutions adapted to a fixed filtration and driven by a given Wiener process. Recent work shows that such solutions exist in abundance, but also that pathwise non-uniqueness can be severe.

For stochastic shear-thinning power-law flows on $L^2_\sigma$4, a new energy related functional is introduced into convex integration, yielding infinitely many Leray–Hopf solutions that are probabilistically strong for a certain initial value. The same work provides global in time estimates leading to the existence of infinitely many stationary and even ergodic Leray–Hopf solutions (Berkemeier, 2024). For the three-dimensional fractional Navier–Stokes equations perturbed by transport noise, there exist infinitely many Hölder continuous analytically weak, probabilistically strong Leray–Hopf solutions starting from the same deterministic initial velocity field; the energy inequality holds pathwise on a non-empty random interval $L^2_\sigma$5 (Lange et al., 2024).

The non-uniqueness results become sharper for stochastic Navier–Stokes equations with linear multiplicative noise. For stochastic forced Navier–Stokes equations on $L^2_\sigma$6, there are two distinct adapted Leray–Hopf solutions on a random interval $L^2_\sigma$7 driven by the same Brownian motion, and there is also joint non-uniqueness in law for global solutions on $L^2_\sigma$8 (Hofmanová et al., 2023). For stochastic fractional Navier–Stokes equations with linear multiplicative noise on $L^2_\sigma$9, non-uniqueness of local strong solutions holds for certain deterministic forces, and for some stochastic force the system admits two different global Leray–Hopf solutions smooth on any compact subset of $\partial_t u,\nabla^2u\in L^2$0 (Chen et al., 23 Mar 2025).

A contrasting phenomenon appears in stochastic hyper-viscous Navier–Stokes beyond the Lions exponent. In that regime, there exist infinitely many probabilistically strong and analytically weak solutions in supercritical mixed spaces $\partial_t u,\nabla^2u\in L^2$1, but these are explicitly non-Leray–Hopf; at the same time, uniqueness still holds in the Leray–Hopf class when $\partial_t u,\nabla^2u\in L^2$2 (Cao et al., 2024). This contrast isolates the energy inequality as a decisive selection principle: outside the Leray–Hopf class, pathwise strong non-uniqueness may persist even in highly dissipative regimes.

6. Right continuity, energy regularization, and immediate strongness

Another line of work emphasizes that strongness may refer not to higher spatial integrability, but to stronger temporal behavior inside the energy class. One abstract gives the notion of energy-regularized solutions of the three-dimensional Navier–Stokes equations, proves that each ER-solution satisfies Leray–Hopf property, and proves that each ER-solution is rightly continuous in the standard phase space $\partial_t u,\nabla^2u\in L^2$3 endowed with the strong convergence topology (Kasyanov, 2023). A related abstract states that each weak solution for the 3D Navier–Stokes system satisfies Leray–Hopf property and is rightly continuous in the standard phase space $\partial_t u,\nabla^2u\in L^2$4 endowed with the strong convergence topology (Gorban et al., 2015).

This suggests a distinct sense of “strong Leray–Hopf”: a solution that remains analytically weak but is strongly right-continuous in $\partial_t u,\nabla^2u\in L^2$5. In that interpretation, the strengthened property is temporal rather than spatial.

A further variant appears in the two-dimensional inhomogeneous Navier–Stokes system. There, one studies Leray–Hopf weak solutions that are immediately strong, meaning that for every $\partial_t u,\nabla^2u\in L^2$6,

$\partial_t u,\nabla^2u\in L^2$7

The corresponding theory proves that the following are equivalent: being immediately strong; satisfying the strong energy inequality; obeying the quantitative bounds $\partial_t u,\nabla^2u\in L^2$8; and admitting an associated pressure

$\partial_t u,\nabla^2u\in L^2$9

As an application, a weak-strong uniqueness theorem is obtained (Crin-Barat et al., 2024).

In this two-dimensional inhomogeneous setting, the “quenched” feature is not probabilistic or Baire-generic; rather, it is the instantaneous transition from a weak energy solution at $L^r(0,T;L^s)$0 to a strong solution for every positive time. The paper explicitly presents this as a way to characterize Leray–Hopf solutions that become strong for positive times and to organize several recent weak-strong uniqueness results into a unified framework.

7. Conceptual synthesis

Taken together, the recent literature supports a stratified view of quenched strong Leray–Hopf solutions. In deterministic three-dimensional Navier–Stokes, the generic Baire-category analysis is mostly negative: for a Baire generic $L^r(0,T;L^s)$1 datum, Leray–Hopf solutions generally fail the known super-Euler-scale integrability conditions that imply uniqueness or energy equality, and the only plausible robust strong theory lies on the Euler-critical line (Lindberg, 2024). In stochastic equations, the pathwise notion is positive at the level of existence—there are many probabilistically strong Leray–Hopf solutions—but often negative at the level of uniqueness, since convex integration and unstable-manifold constructions can produce multiple pathwise Leray–Hopf solutions driven by the same noise realization (Berkemeier, 2024, Chen et al., 23 Mar 2025).

Across both settings, the decisive role is played by selection principles: critical scaling, viscosity-uniform a priori estimates, the energy inequality, energy equality, right continuity in the strong $L^r(0,T;L^s)$2 topology, or immediate strongness for positive times. A plausible implication is that “quenched strong Leray–Hopf solution” is best understood not as a single established object, but as a family of strengthened Leray–Hopf notions tailored to different problems: Baire-generic strongness in deterministic theory, pathwise strong admissibility in stochastic theory, and instant positive-time regularization in inhomogeneous two-dimensional flows.

The unifying theme is that the Leray–Hopf class by itself is usually too broad to enforce strong behavior, while genuinely quenched strong behavior—whether generic or pathwise—appears only under additional rigidity: critical scaling, stronger energy structure, or explicit regularity assumptions.