🔢 mathematics

Probability density functions as solutions of heterogeneous Cattaneo-Vernotte diffusion equation

This paper derives exact analytical solutions for a heterogeneous Cattaneo-Vernotte diffusion equation with an exponential diffusion coefficient in terms of modified Bessel functions and proves that these solutions constitute valid probability density functions.

Original authors: K. Górska, Ž. Tomovski, A. Horzela, M. Magdziarz

Published 2026-06-30

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: K. Górska, Ž. Tomovski, A. Horzela, M. Magdziarz

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to track a drop of ink spreading through a sponge. In a normal, uniform sponge, the ink spreads out smoothly and predictably. But what if the sponge is "heterogeneous"? That means some parts are dense and hard to move through, while other parts are loose and easy to move through. The ink doesn't just spread; it gets stuck, speeds up, and changes its behavior depending on exactly where it is.

This paper is about solving a very specific mathematical puzzle regarding how things spread (diffuse) in these uneven, "sponge-like" environments. Here is the breakdown of what the authors did, using simple analogies.

The Problem: The "Lazy" Diffusion

Usually, scientists use a standard equation to describe how things spread. But in complex materials, things don't react instantly. There is a tiny "lag" or a moment of hesitation before the spread happens. The authors used a more advanced equation called the Cattaneo-Vernotte equation to account for this delay.

They also added a twist: the "resistance" of the material changes exponentially as you move through it (like a sponge that gets progressively denser the further you go).

The Solution: A Mathematical Recipe

The authors wanted to find the exact formula that describes the position of our "ink drop" at any given time. Instead of just guessing, they used a powerful mathematical tool called the Laplace transform. Think of this as a translator that turns a complicated, moving puzzle into a static algebra problem that is easier to solve.

After doing the heavy lifting, they found two exact solutions. These solutions are written as a ratio of special mathematical shapes (called Modified Bessel functions).

Analogy: Imagine trying to describe the shape of a shadow cast by a complex 3D object. Instead of describing the shadow directly, they found a way to describe it by comparing two specific, well-known geometric curves.

The Big Question: Is it a Real Probability?

In physics, when we talk about the position of a particle, we use a Probability Density Function (PDF). This is just a fancy way of saying a map that tells us the chance of finding the particle at a specific spot.

Rule 1: The chance must never be negative (you can't have a -50% chance of finding something).
Rule 2: If you add up all the chances across the whole universe, it must equal 100% (the particle has to be somewhere).

The authors had to prove that their new mathematical formulas actually obeyed these rules. They couldn't just say "it looks right"; they had to prove it mathematically.

The Proof: The "Monotonicity" Test

To prove their formulas were valid probability maps, they used a concept called Complete Monotonicity.

The Metaphor: Imagine a hill that is perfectly smooth and always slopes downward. No bumps, no dips, no going back up. If a mathematical function behaves like this "perfectly sloping hill," it is guaranteed to represent a valid, non-negative probability.
The authors showed that their complex Bessel function ratios behave exactly like this "perfectly sloping hill." They proved that no matter how you look at the math, the result is always a positive, valid probability.

The Two Scenarios

The paper found that the behavior of the "ink" depends on a specific parameter (let's call it the "stochastic knob," denoted by $\alpha$ ).

Scenario A: If the knob is set between 0.5 and 1, the solution uses one type of Bessel function ratio (involving the $I$ function).
Scenario B: If the knob is set between 0 and 0.5, the solution uses a different type of Bessel function ratio (involving the $K$ function).

In both cases, they proved that the math works out to be a valid probability map.

The Conclusion

The authors successfully:

Took a complex equation describing diffusion in uneven, changing materials with a time delay.
Solved it to find exact formulas involving special mathematical curves.
Proved rigorously that these formulas represent real, physical probabilities (they are never negative and always sum to 100%).

They did this entirely through mathematical theory, ensuring that the "ink drop" in their model behaves logically and physically, even in the most complex, uneven environments.

Technical Summary: Probability Density Functions as Solutions of Heterogeneous Cattaneo-Vernotte Diffusion Equation

Problem Statement
The paper addresses the mathematical characterization of diffusion processes in heterogeneous media, specifically focusing on the Cattaneo-Vernotte (CV) equation. Unlike standard diffusion models, the CV equation incorporates a time lag ( $\tau$ ) to account for finite propagation speeds, making it suitable for modeling phenomena where the diffusive flux does not respond instantaneously to concentration gradients. The authors investigate a specific heterogeneous case where the diffusion coefficient $D(x)$ varies exponentially with position, defined as $D_\lambda(x) = B\exp(\lambda|x|)$ .

The core challenge is to determine whether the solutions to this equation, subject to fundamental initial conditions (a Dirac delta initial distribution and zero initial time derivative) and boundary conditions (vanishing at infinity), constitute valid probability density functions (PDFs). Specifically, the authors aim to prove the non-negativity and normalization of these solutions, which are prerequisites for a physical interpretation as probability distributions.

Methodology
The authors employ a combination of integral transform techniques and the theory of completely monotonic (CM) functions to solve and analyze the equation:

Laplace Transform: The heterogeneous CV equation is transformed into the Laplace domain (variable $s$ ). Due to the symmetry of the diffusion coefficient $D_\lambda(x)$ , the problem is reduced by introducing a coordinate transformation $u = \exp(-|x|)$ .
Bessel Differential Equations: The transformed equation in the $u$ -coordinate is identified as a Bessel-type differential equation. The authors solve this to obtain exact analytical expressions for the Laplace-transformed solution $\hat{p}(x,s)$ . These solutions are expressed as ratios of modified Bessel functions ( $I_\nu$ and $K_\nu$ ) with arguments involving $\sqrt{\tau s^2 + s}$ .
Complete Monotonicity Analysis: To establish non-negativity in the time domain, the authors utilize Bernstein's theorem, which states that a function is non-negative if its Laplace transform is a completely monotonic (CM) function. They decompose the Laplace solutions into products of functions and apply properties of CM and Bernstein (B) functions:
- Property 1: The product of two CM functions is CM.
- Property 2: The composition of a CM function with a B function is CM.
- Property 3: Linear combinations of CM functions on a convex cone are CM.
Theoretical Proofs: The paper provides rigorous proofs (Theorems 1–4) demonstrating that specific ratios of modified Bessel functions, such as $I_{\alpha-1}(s)/I_\alpha(s)$ and $K_{\alpha-1}(s)/K_\alpha(s)$ , are CM functions within specific ranges of the parameter $\alpha \in [0, 1]$ . These proofs rely on integral representations involving Jacobi theta functions and spectral functions derived from the zeros of Bessel functions.

Key Contributions and Results
The paper derives two distinct exact analytical solutions for the heterogeneous CV equation, corresponding to the sign of the parameter $\lambda$ :

Case 1 ( $\lambda > 0$ ): The solution involves the ratio of modified Bessel functions of the first kind, $I_\nu$ . The authors prove that the inverse Laplace transform of this solution is a valid PDF for the parameter range $\alpha \in [1/2, 1)$ .
Case 2 ( $\lambda < 0$ ): The solution involves the ratio of modified Bessel functions of the second kind, $K_\nu$ . The authors prove that this solution corresponds to a valid PDF for the parameter range $\alpha \in [0, 1/2]$ .

The solutions are presented in terms of the ratio of modified Bessel functions with arguments scaled by $\sqrt{\tau s^2 + s}$ , reflecting the interplay between the first and second time derivatives inherent in the CV equation.

Furthermore, the authors demonstrate that these solutions are normalized with respect to the spatial variable $x \in \mathbb{R}$ , confirming that the total probability is conserved.

Significance and Claims
The paper claims to generalize previous mathematical results regarding the complete monotonicity of ratios of Bessel functions (specifically those related to the Student t-distribution) to the context of the heterogeneous Cattaneo-Vernotte equation. By introducing position-dependent parameters $a(x)$ and $b(x)$ and modifying the arguments of the Bessel functions to include the term $\sqrt{\tau s^2 + s}$ , the authors establish a direct link between the mathematical structure of the equation and the physical requirement of non-negativity for probability densities.

The primary significance lies in the rigorous proof that the solutions to this specific class of heterogeneous diffusion equations are indeed probability density functions. This validates the use of the Cattaneo-Vernotte model with exponential diffusion coefficients for describing stochastic processes in heterogeneous media, ensuring that the resulting distributions are physically meaningful (non-negative and normalized). The work relies on the theory of completely monotonic functions to bypass the difficulty of directly inverting the complex Laplace transforms, instead proving the necessary properties in the transform domain.