pith. sign in

arxiv: 2604.02519 · v1 · submitted 2026-04-02 · ⚛️ physics.ao-ph · math.PR

On the White-Noise Limit of the Colored Linear Inverse Model

Cristian Martinez-Villalobos This is my paper

Pith reviewed 2026-05-13 20:22 UTC · model grok-4.3

classification ⚛️ physics.ao-ph math.PR
keywords colored noiselinear inverse modelwhite-noise limitOrnstein-Uhlenbeck processfluctuation-dissipation relationstochastic differential equationscorrelation time
0
0 comments X

The pith

As noise correlation time approaches zero, the colored linear inverse model reduces to the classical white-noise LIM and satisfies the fluctuation-dissipation relation.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper shows that representing the colored LIM as an augmented Ornstein-Uhlenbeck system allows the white-noise limit to be taken directly on the stochastic differential equations. As the correlation time tau tends to zero, the colored-noise-driven system converges to the classical LIM. In this limit the stationary covariance obeys the standard fluctuation-dissipation relation. This regular convergence of the dynamics stands in contrast to the singular behavior of derivative-based identification formulas. Numerical checks on the same linear system used in prior work confirm the convergence.

Core claim

Treating the colored LIM as an augmented Ornstein-Uhlenbeck system, we show that as the correlation time tau -> 0 the colored-noise-driven system reduces to the classical LIM, and the corresponding stationary covariance satisfies the standard fluctuation-dissipation relation. Re-examining the same linear system used by Lien et al. (2025), we illustrate this convergence numerically.

What carries the argument

Augmented linear Ornstein-Uhlenbeck system that embeds the colored noise into a larger white-noise-driven linear model

If this is right

  • The stationary covariance of the colored LIM converges to that of the classical LIM.
  • The fluctuation-dissipation relation holds exactly in the white-noise limit.
  • The underlying stochastic dynamics converge regularly despite the ill-defined nature of derivative-based estimation formulas.
  • Convergence of estimated parameters obtained by other methods is consistent with this dynamical limit.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Colored-noise formulations may serve as a regularization device that recovers white-noise models in a controlled limit.
  • Identification procedures for LIMs must be selected according to whether the driving noise is treated as colored or white to avoid singularities.
  • The same augmented-system approach could be tested on nonlinear inverse models to check whether analogous regular limits exist.

Load-bearing premise

The colored LIM can be exactly represented as an augmented linear Ornstein-Uhlenbeck system whose white-noise limit is taken directly on the stochastic differential equations.

What would settle it

A numerical computation of the stationary covariance matrix for small but nonzero tau that deviates from the fluctuation-dissipation relation predicted for the classical LIM would falsify the claimed convergence.

Figures

Figures reproduced from arXiv: 2604.02519 by Cristian Martinez-Villalobos.

Figure 1
Figure 1. Figure 1: FIG. 1. Convergence test using the system of [ [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
read the original abstract

A recent paper by Lien et al. (2025) introduces the "colored linear inverse model" (colored LIM), in which stochastic forcing is modeled using Ornstein-Uhlenbeck colored noise rather than idealized white noise. In that work, it is shown that the derivative-based identification formulas used to estimate model parameters do not admit a regular white-noise limit due to the loss of differentiability of the lag-correlation function at zero lag. Here we revisit the white-noise limit from the perspective of the underlying stochastic differential equations. Treating the colored LIM as an augmented Ornstein-Uhlenbeck system, we show that as the correlation time tau -> 0 the colored-noise-driven system reduces to the classical LIM, and the corresponding stationary covariance satisfies the standard fluctuation-dissipation relation. Re-examining the same linear system used by Lien et al. (2025), we illustrate this convergence numerically. These results highlight a distinction between the singular behavior of derivative-based identification formulas and the regular limiting behavior of the underlying stochastic model. Taken together with recent results showing convergence of estimated parameters in the white-noise limit, they provide a consistent interpretation in which the colored LIM recovers the classical LIM at the level of stochastic dynamics even though certain estimation procedures become ill-defined in that limit.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript shows that the colored linear inverse model (LIM), formulated as an augmented linear Ornstein-Uhlenbeck process with colored noise, converges in the white-noise limit (correlation time τ → 0 with noise amplitude scaled to keep integrated power finite) to the classical white-noise LIM. The stationary covariance of the projected system satisfies the standard fluctuation-dissipation relation expressed by the Lyapunov equation, and this convergence is illustrated numerically on the linear system examined by Lien et al. (2025).

Significance. If the derivation holds, the result clarifies that the underlying stochastic dynamics of the colored LIM recover the classical LIM regularly in the white-noise limit, in contrast to the singular behavior of derivative-based identification formulas. This provides a consistent dynamical interpretation and complements recent findings on parameter convergence. The augmented-SDE representation and numerical check are positive features, relying on standard continuity properties of stationary covariances for linear SDEs.

major comments (2)
  1. [§2] §2 (augmented SDE representation): the exact scaling of the noise intensity with τ (to hold integrated power finite) must be stated explicitly, including the resulting limiting diffusion matrix, to confirm direct recovery of the classical LIM drift and diffusion terms.
  2. [§3] §3 (stationary covariance limit): the argument that the projected stationary covariance satisfies the standard FDR (Lyapunov equation) of the limiting system should include a brief justification that the projection commutes with the limit, or cite the relevant continuity result for the covariance map.
minor comments (2)
  1. [Numerical example] In the numerical section, report quantitative measures (e.g., Frobenius norm of the difference from the Lyapunov solution) for the smallest τ values to make the convergence rate visible.
  2. [References] Ensure the reference list includes the full bibliographic details for Lien et al. (2025) and any theorems on OU-process convergence invoked in the proof.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their positive assessment and recommendation for minor revision. We address the two major comments below and will incorporate the requested clarifications.

read point-by-point responses

  1. Referee: [§2] §2 (augmented SDE representation): the exact scaling of the noise intensity with τ (to hold integrated power finite) must be stated explicitly, including the resulting limiting diffusion matrix, to confirm direct recovery of the classical LIM drift and diffusion terms.

    Authors: We agree that the scaling should be stated explicitly. In the revised manuscript we will add the precise scaling: the Ornstein–Uhlenbeck forcing η satisfies dη = −η/τ dt + √(2D/τ) dW, so that its integrated power remains finite and equal to D as τ → 0. We will then show that the augmented (x, η) system converges in the appropriate sense to the classical LIM dx = Ax dt + √D dW, with the limiting diffusion matrix recovered directly as D. This makes the recovery of both drift and diffusion terms explicit. revision: yes

  2. Referee: [§3] §3 (stationary covariance limit): the argument that the projected stationary covariance satisfies the standard FDR (Lyapunov equation) of the limiting system should include a brief justification that the projection commutes with the limit, or cite the relevant continuity result for the covariance map.

    Authors: We will add the requested justification. The stationary covariance of a linear SDE is a continuous function of the drift and diffusion matrices (via the unique positive-definite solution of the Lyapunov equation, which depends continuously on the coefficients when the drift is Hurwitz). Because the augmented-system coefficients converge to those of the classical LIM as τ → 0, the stationary covariance converges to the classical one. The subsequent projection onto the x-variables is a fixed linear map and therefore commutes with the limit. A short paragraph citing this standard continuity property of the Lyapunov solution will be inserted in §3. revision: yes

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper derives the white-noise limit by representing the colored LIM exactly as an augmented linear Ornstein-Uhlenbeck system and taking the tau -> 0 limit directly on the underlying SDEs, with noise amplitude scaled to preserve integrated power. The stationary covariance of the limiting system then satisfies the classical Lyapunov equation (fluctuation-dissipation relation) by continuity of the covariance map for linear SDEs. This is a standard convergence result for OU processes and does not reduce any claimed prediction to a fitted quantity defined by the paper itself, nor does it invoke load-bearing self-citations or ansatzes smuggled from prior work. The numerical illustration on the Lien et al. example is consistent with but not required for the analytic argument. The derivation is therefore self-contained.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The argument rests on standard properties of linear Ornstein-Uhlenbeck processes and the fluctuation-dissipation theorem for white-noise driven linear systems; no new free parameters or invented entities are introduced.

axioms (2)
  • domain assumption Colored noise is exactly represented by an Ornstein-Uhlenbeck process
    Standard modeling choice for finite-correlation-time forcing in linear inverse models.
  • domain assumption The system remains linear after augmentation
    Invoked when treating the colored LIM as an augmented Ornstein-Uhlenbeck system.

pith-pipeline@v0.9.0 · 5518 in / 1299 out tokens · 50274 ms · 2026-05-13T20:22:07.138416+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

10 extracted references · 10 canonical work pages

  1. [1]

    Lien, Y.-N

    J. Lien, Y.-N. Kuo, H. Ando, and S. Kido, Colored linear inverse model: A data-driven method for studying dynamical systems with temporally correlated stochasticity, Physical Review Research7, 023042 (2025)

    work page 2025

  2. [2]

    J. Lien, H. Ando, I. Richter, and S. Kido, Noise Persistence in a Cyclostationary Linear Inverse Model Framework and Its Impact on ENSO Evolution and Prediction Skills, Journal of the Atmospheric Sciences83, 391 (2026)

    work page 2026

  3. [3]

    Penland and P

    C. Penland and P. D. Sardeshmukh, The Optimal Growth of Tropical Sea Surface Temperature Anomalies, Journal of Climate8, 1999 (1995)

    work page 1999

  4. [4]

    Gushchina and B

    D. Gushchina and B. Dewitte, Intraseasonal Tropical Atmospheric Variability Associated with the Two Flavors of El Ni˜ no, Monthly Weather Review140, 3669 (2012)

    work page 2012

  5. [5]

    Thual, A

    S. Thual, A. J. Majda, N. Chen, and S. N. Stechmann, Simple stochastic model for El Ni˜ no with westerly wind bursts, Proceedings of the National Academy of Sciences113, 10245 (2016)

    work page 2016

  6. [6]

    Martinez-Villalobos, M

    C. Martinez-Villalobos, M. Newman, D. J. Vimont, C. Penland, and J. David Neelin, Observed El Ni˜ no-La Ni˜ na Asymmetry in a Linear Model, Geophysical Research Letters46, 9909 (2019)

    work page 2019

  7. [7]

    P. Sura, M. Newman, and M. A. Alexander, Daily to Decadal Sea Surface Temperature Vari- ability Driven by State-Dependent Stochastic Heat Fluxes, Journal of Physical Oceanography 36, 1940 (2006)

    work page 1940

  8. [8]

    N. G. Van Kampen, Chapter IX - THE LANGEVIN APPROACH, inStochastic Processes in Physics and Chemistry (Third Edition), North-Holland Personal Library, edited by N. G. Van kampen (Elsevier, Amsterdam, 2007) pp. 219–243

    work page 2007

  9. [9]

    Gardiner,Stochastic Methods(2009)

    C. Gardiner,Stochastic Methods(2009)

    work page 2009

  10. [10]

    Penland and L

    C. Penland and L. Matrosova, A Balance Condition for Stochastic Numerical Models with Application to the El Ni˜ no-Southern Oscillation, Journal of Climate7, 1352 (1994). 8

    work page 1994