Modulus-Dependent Random Renewing Flow

• The paper demonstrates that modulus-dependent flows enable spectral tuning by continuously varying the elliptic modulus to control Lyapunov exponents and cumulants.
• It details the construction and analysis using perturbative expansions about the angular Laplacian, offering explicit series for exponents in d=2 and d=3.
• The study reveals practical implications in passive scalar transport where ergodicity and deterministic effective diffusivity emerge from stochastic advection models.

A modulus-dependent random renewing flow is a class of stochastic velocity fields in which the statistical and spectral properties of the flow, and related physical observables, depend continuously on a modulus parameter (typically an elliptic modulus $k\in[0,1)$). Such models generalize classical random renewing flows, allowing spectral tuning and finer probing of ergodic and dispersion phenomena. These flows are studied in the context of advection–diffusion of passive scalars, random matrix products arising from linearization, and continuum limits describing the long-term statistical evolution of transported quantities (Ding et al., 2021, Tourigny, 23 Nov 2025). The modulus parameter controls the spectral perturbation away from angular Laplacians, with key consequences for Lyapunov exponents, cumulants, and effective diffusivity tensors.

1. Definition and Canonical Construction

In the prototypical modulus-dependent random renewing flow model, the $d$-dimensional velocity field $u(x, t)$ is defined as a divergence-free, piecewise-constant-in-time random process. At each renewal time, the entire velocity realization is replaced by an independent sample. The temporal correlation is thus restricted to intervals of fixed duration, leading to block-independent statistics: $$u_i(x, t) = \sum_{j\neq i} u_{ij}(x_j + \eta_{ij}(t)), \quad i=1,\dots,d,$$ where $u_{ij}$ are $2\pi$-periodic, zero-mean functions of one variable, and $\eta_{ij}(t)$ are independent random shifts renewed at each block.

Linearization yields a discrete Jacobian map $\delta x^{n+1} = g_n\,\delta x^n$ where $g_n \in \mathrm{SL}(d, \mathbb{R})$ factors into exponentials of random Lie algebra generators, with randomness controlled by the ensemble of $u'_{ij}$ evaluated at random phases. The ensemble is characterized by a continuum-limit parameter $\tau = \sigma / (\sqrt{2}\,\delta t)$, with $\sigma^2 = \mathbb{E}[(u'_{ij})^2]$.

The modulus dependence arises in the spectral analysis of the associated transfer operators, where the perturbation parameter $k^2$ (the square of the elliptic modulus) continuously interpolates between the pure Laplacian (angular diffusion) and the fully “symmetric disorder” regime ($k^2=1/d$ for $d$ dimensions) (Tourigny, 23 Nov 2025).

2. Modulus-Dependence in Spectral Problems

The stochastic flow induces random matrix products whose statistical properties (e.g., asymptotic expansion rates) are encoded in generalized Lyapunov exponents. In the continuum (Kraichnan–Kazantsev) limit, these problems reduce to spectral theory for second-order partial differential operators with explicit modulus dependence: $$\mu\,v = \bigl[\,\Delta_K - k^2\sum_{i<j} A_{ij}^2\,\bigr] v,$$ where $\Delta_K$ is the Laplacian (Casimir) on the compact subgroup $K = \mathrm{SO}(d)$ and $A_{ij}$ the Cartan generators in $\mathfrak{sl}(d)$. The modulus $k$ governs the coefficient of the non-compact part of the operator.

In $d=2$, the angular parameterization reduces the problem to an ODE in $\theta$, with spectral perturbation involving $k^2 \sin^2(2\theta)$. In $d=3$, the problem is formulated on the sphere $S^2$ with spherical harmonics as the unperturbed eigenbasis, and the perturbation $-k^2\sum A_{ij}^2$ is block-diagonal in $l$.

This setting allows for perturbative expansions in powers of $k^2$ about the angular-Laplacian (modulus $k=0$) limit, yielding explicit series for exponents and cumulants (Tourigny, 23 Nov 2025).

3. Effective Diffusivity and Ergodicity

In the context of passive scalar transport, modulus-dependent random renewing flows generate a deterministic effective diffusivity tensor in the long-time limit. The mean scalar field $\overline{T}(x, t)$ evolves, after center-manifold reduction, according to

$$\partial_t \overline{T} = D_{\mathrm{eff}}\,\partial_x^2 \overline{T},$$

with

$$D_{\mathrm{eff}} = 1 + \mathrm{Pe}^2 \sigma_A^2\,\frac{1}{L_t} \sum_{k=1}^\infty \langle u, \phi_k\rangle^2 \int_0^{L_t}\int_0^s e^{\lambda_k(\sigma-s)}\xi(s)\xi(\sigma)\,d\sigma\,ds.$$

Here, $\sigma_A^2$ is the variance of the random amplitudes, $\xi$ the periodic temporal profile, and $\{\phi_k\}$ the Neumann eigenbasis of the cross-sectional Laplacian.

By the renewal structure (i.i.d.~random amplitudes) and Weyl mixing of $\xi$, one obtains a law of large numbers effect: for almost all single realizations, the streamwise variance of the scalar converges to the deterministic $D_{\mathrm{eff}}$. This establishes a genuine ergodicity result: deterministic long-time behavior emerges despite persistent randomness in the flow (Ding et al., 2021).

A noteworthy result is that $$D_{\mathrm{eff}}^{\text{random}} \ge D_{\mathrm{eff}}^{\text{deterministic}}$$, with equality for the leading blockwise integral, and strict inequality due to small cross-block corrections which always lower the effective diffusivity.

4. Expansion of Generalized Lyapunov Exponents

In the spectral theory of random matrix products associated with modulus-dependent flows, the generalized Lyapunov exponent $L(k, \ell)$ admits an explicit power series in $k^2$: $$L(k, \ell) = \tau^2\sum_{n=0}^\infty a_n(\ell)\,k^{2n},$$ where $\tau$ is the continuum-limit parameter.

For $d=2$, the coefficients are

$$a_0(\ell)=\frac{\ell(\ell+2)}{2}, \quad a_1(\ell)=-\frac{\ell(\ell+2)}{2}, \quad a_2(\ell)=\frac{(\ell-2)\ell(\ell+2)(\ell+4)}{128}, \ldots$$

For $d=3$: $$\frac{1}{\tau^2}L(k, \ell) = \frac{2\ell(\ell+3)}{3} - \frac{4\ell(\ell+3)}{5}k^2 + \frac{12(\ell-2)\ell(\ell+3)(\ell+5)}{875}k^4 - \frac{72(\ell-2)\ell(\ell+3)(\ell+5)(9\ell^2+27\ell-395)}{3\,128\,125}k^6 + \cdots$$ These expansions allow for systematic calculation of exponents and higher-order cumulants at small modulus (Tourigny, 23 Nov 2025).

At fixed $k$, the first two cumulants in $d=2$ are: $$\frac{\gamma_1(k)}{\tau^2} = 2\Bigl(\frac{E}{K}-1\Bigr), \quad \frac{\gamma_2(k)}{2\tau^2} = \frac{3}{2} - k^2 - \frac{E}{K} + \cdots,$$ where $K=K(k)$ and $E=E(k)$ are the complete elliptic integrals.

5. Special Regimes and Recovery of Classical Limits

In the special case where the temporal profile $\xi(t)$ is constant and the modulus $k\mapsto0$, the modulus-dependent random renewing flow reduces to the classical Taylor–Aris setup for shear dispersion, with: $$D_{\mathrm{eff}} = 1 + \mathrm{Pe}^2 (\overline{u^2} - \overline{u}^2).$$ Likewise, the full leading Lyapunov exponent for the canonical choice $k^2=1/d$ matches established results for isotropic (symmetric disorder) Kraichnan models.

For $d=2$, $k^2=1/2$ yields the classical Lyapunov exponent

$$\frac{\gamma_1(k)}{\tau^2} = 2\bigl(E/K - 1\bigr) \approx 0.3513,$$

as reported in fluid mechanics literature (Tourigny, 23 Nov 2025).

6. Relation to Angular Laplacians, Eigenfunctions, and Perturbation Theory

The modulus-dependent problem is naturally represented as a perturbation of the angular Laplacian on spheres. In $d=2$, the base operator is the Laplacian on $S^1$; in $d=3$, on $S^2$ with spherical harmonics $Y_l^m$ as eigenfunctions. The perturbative $k^2$ term retains block-diagonality under these bases, enabling calculation of spectral corrections via finite-dimensional perturbation theory.

The modulus thereby encodes a continuous interpolation between pure rotational diffusion and regimes with enhanced stretching or mixing due to the non-compact generator components. In all cases, the analytic structure of the expansions is governed by the underlying group-theoretic symmetries (Tourigny, 23 Nov 2025).

7. Significance and Applications

Modulus-dependent random renewing flows provide a unifying framework for analyzing random advection, dispersion, and stretching in both deterministic and stochastic shear environments, with precise control of spectral properties via the modulus parameter. This framework is applicable to problems in passive scalar mixing, fluid transport, and the statistical mechanics of random matrix products. The ergodic and deterministic long-time behavior, explicit series for exponents, and recoverability of known limits underpin its foundational role in modern homogenization theory and random dynamical systems (Ding et al., 2021, Tourigny, 23 Nov 2025).

A plausible implication is that modulus-dependent models offer flexible testbeds for probing the transition between deterministic and random transport, guiding experimental and computational studies where tunable disorder is intrinsic to the system.