Late-Time Relaxation from Landau Singularities

Imagine you are watching a cup of hot coffee cool down on a table. At first, the steam rises vigorously, and the temperature drops quickly. This is the "early time" behavior, where the specific details of the coffee molecules matter a lot. But as time goes on, the coffee settles into a slow, steady decline toward room temperature. This is the "late time" behavior.

For a long time, scientists thought this slow decline always followed a simple, predictable rule: it would drop off like a ball bouncing on a trampoline, getting smaller and smaller at a steady, exponential rate (like $e^{-t}$ ).

However, this paper argues that in many real-world systems, the story is more like a slowly fading echo than a bouncing ball. Instead of dropping off quickly, the system's fluctuations (tiny jitters in temperature, pressure, or density) linger much longer, decaying according to a "power law" (like $1/t$ ). This means they hang around for a very long time, much slower than previously thought.

Here is how the authors figured this out, using simple analogies:

1. The Crowd and the Whisper (Fluctuations)

In any large system (like a gas, a fluid, or even the early universe), particles are constantly jiggling around due to heat. These jiggles are called fluctuations.

The Old View: Scientists used to think these jiggles were just background noise, like static on a radio, that could be ignored or treated as independent whispers.
The New View: The authors show that these whispers actually talk to each other. When one particle jiggles, it bumps into its neighbors, which then bump into others. These nonlinear interactions create a chain reaction.

2. The "Banana" Shape (The Mathematical Tool)

To understand how these whispers interact, the authors use a framework called Schwinger-Keldysh Effective Field Theory. Think of this as a sophisticated rulebook for tracking how energy and noise move through a system.

In this rulebook, the interactions between particles are drawn as diagrams. The most important shape here is called a "banana diagram."

Imagine a banana. It has two ends (the start and end of a process) and a curved body in the middle.
In the math, this shape represents a particle going out, interacting with the "soup" of other particles (the loop in the middle), and coming back.
The authors realized that to find out how long the system takes to relax, you don't need to do the incredibly hard math of calculating every single bump in the loop. Instead, you just need to look at the shape of the banana.

3. The Landau Singularity (The Pinch Point)

The core of the paper is a technique called Landau singularity analysis.

The Analogy: Imagine you are walking through a crowded market. Usually, you can walk freely. But at a specific moment, the crowd squeezes together so tightly from both sides that you get "pinched" and can't move forward or backward. That pinch point is a singularity.
In the math of these particle loops, a "pinch" happens when the paths of different particles align perfectly. The authors used a set of algebraic rules (the Landau equations) to find exactly where these pinch points happen without doing the heavy lifting of the full calculation.

4. The Result: The "Gapless" Echo

When the authors analyzed these pinch points, they found something surprising:

If the system has "gapless" modes (meaning there are no barriers stopping the fluctuations, like sound waves in air or heat in a fluid), the "pinch" creates a new kind of decay.
Instead of the fast, exponential drop-off (the bouncing ball), the system enters a power-law decay.
The Metaphor: Think of a bell. If you hit it, it rings loudly and then fades away quickly (exponential). But if you have a system with these specific nonlinear interactions, it's more like a bell in a canyon. The sound bounces off the walls, creating a long, lingering echo that fades very slowly. The "power law" is the mathematical description of that lingering echo.

Summary of the Discovery

The paper provides a systematic way to predict this "lingering echo" in almost any macroscopic system (like fluids or heat conductors) without needing to solve complex integrals.

The Claim: Nonlinear interactions (particles bumping into each other) create new "decay modes" that are much slower than the basic ones.
The Mechanism: These slow modes are caused by "pinch points" (Landau singularities) in the mathematical description of particle loops (banana diagrams).
The Outcome: When these slow modes exist, the system's relaxation at late times follows a power law ( $1/t$ ) rather than an exponential curve.

The authors emphasize that this is a universal feature of systems with conservation laws (like conservation of energy or momentum) and nonlinear interactions. It explains why things in the real world often take much longer to settle down than simple linear models predict.

Technical Summary: Late-Time Relaxation from Landau Singularities

Problem Statement
Macroscopic fluctuations in finite-temperature systems, while often negligible in relative magnitude, can control observable phenomena in specific regimes such as relativistic heavy-ion collisions, two-dimensional systems with continuous symmetry, and gravitational-wave detectors. A central question in statistical physics is how these fluctuations relax. While early-time dynamics depend on microscopic details, late-time relaxation is governed by macroscopic symmetries, conservation laws, and constitutive relations.

In linear response theory, fluctuation eigenmodes decay exponentially. However, nonlinear hydrodynamic interactions couple these modes, invalidating the independent-mode picture and generating singularities absent from the linear spectrum. These nonlinear interactions can replace exponential relaxation with "long-time tails," characterized by power-law decay in correlation functions. While this phenomenon has been identified in specific models (e.g., velocity autocorrelation functions) and calculated explicitly in nonlinear diffusion models at one- and two-loop orders using the Schwinger-Keldysh effective field theory (SK-EFT), a general characterization of nonlinear late-time relaxation remains elusive. The primary challenge is that late-time behavior is controlled by the singularities of loop integrals, whose structure becomes increasingly intricate beyond simple models or low-loop orders, making explicit integration difficult.

Methodology
This work applies Landau singularity analysis to loop corrections within the SK-EFT framework to determine late-time relaxation behavior without performing explicit loop integrations.

Framework: The authors utilize the SK-EFT formalism, where dynamical fields are decomposed into $r$ (retarded/classical) and $a$ (advanced/stochastic) variables. The effective Lagrangian includes linear terms encoding classical equations of motion and nonlinear terms ( $L_{int}$ ) encoding noise and interactions.
Perturbative Expansion: The full two-point correlation function $G$ is expanded perturbatively in terms of the leading-order propagator $G^{(0)}$ and the self-energy $\Sigma$ . The late-time behavior is governed by the singularities of $G^{(0)}$ and $\Sigma$ in momentum space.
Landau Equations: Instead of evaluating loop integrals directly, the authors employ the Landau equations to identify the singularities of the self-energy $\Sigma$ $Σ$ . These equations determine the conditions under which poles of the propagators in a loop integral pinch the integration contour.
- The analysis focuses on "banana diagrams" (one-particle-irreducible diagrams with $V=2$ vertices and $L$ loops), which are the simplest non-vanishing topologies for $\Sigma_{ar}$ .
- By expanding the basic decay modes at small momentum ( $w(p) \approx -i\gamma - c|p| - i\Gamma|p|^2$ ) and solving the Landau equations, the authors derive the locations of the leading singularities.
Asymptotic Analysis: The singular behavior of the loop integrals near these Landau singularities in the frequency domain is determined. An inverse Fourier transform is then performed to map these frequency-space singularities to the time domain, revealing the asymptotic decay behavior.

Key Contributions and Results

Systematic Identification of Singularities: The paper provides a method to identify the singularities induced by nonlinear interactions directly through algebraic conditions (Landau equations), bypassing the need for explicit loop integration.
General Form of Leading Singularities: For banana diagrams, the authors derive the general form of the leading singularities in the frequency domain:
$\omega^{(s)} = -i\gamma^{(s)} - c|k| - i\Gamma^{(s)}|k|^2$
where $\gamma^{(s)}$ and $\Gamma^{(s)}$ are sums of the parameters of the selected poles from the internal propagators. This result is independent of the specific loop order, noise type, or detailed form of nonlinear interactions.
Emergence of Power-Law Decay:
- If the system possesses gapped modes (where $\text{Re}(\gamma_n) > 0$ ), the new modes generated by nonlinear interactions decay faster than the basic modes, and late-time behavior remains exponential.
- If gapless modes exist (where $\gamma_n = 0$ due to conservation laws), the nonlinear interactions generate new modes with $\gamma^{(s)} = 0$ . These modes decay as $e^{-t\Gamma^{(s)}|k|^2}$ , which is slower than the basic modes and dominates the late-time dynamics.
Universal Power-Law Scaling: In the long-wavelength limit ( $k \to 0$ ), the correlation function exhibits power-law decay:
$\lim_{k \to 0} G(t, k) \approx \text{const} + D(0)t^{-L_{\min}d/2}$
Here, $d$ is the spatial dimension, and $L_{\min}$ is the minimum number of loops required to construct a non-vanishing banana diagram. $L_{\min}$ is determined by the lowest nonlinear power in the equation of motion (e.g., $L_{\min}=1$ for cubic interactions $\phi_a \phi_r^2$ ).
Singular Structure: The analysis shows that the singularities of the self-energy can be either poles or branch points, depending on the integer $\sigma = Ld - V + \kappa$ . For spatial dimensions $d \geq 2$ , these are typically branch points, leading to the power-law behavior.

Significance and Claims
The paper claims to provide a systematic singularity-based description of nonlinear late-time relaxation applicable to a broad class of macroscopic effective theories. By linking Landau singularities in frequency space directly to power-law relaxation in the time domain, the work clarifies the mechanism by which nonlinear mode coupling generates long-time tails.

The authors emphasize that their approach is complementary to previous explicit calculations:

It does not require the explicit construction of the nonlinear part of the operator or the evaluation of complex integrals.
It applies directly to quantities obeying nonlinear equations that are difficult to solve explicitly.
It offers a unified framework to understand the emergence of hydrodynamics and critical slowing down in quantum many-body systems.

The results are presented as a general characterization of how frequency-space singularities control asymptotic time behavior, particularly in systems with gapless modes and nonlinear interactions. The paper does not propose new experimental setups but rather offers a theoretical tool for analyzing existing and future macroscopic effective theories.