Density of reflection resonances in one-dimensional disordered Schrödinger operators

This paper develops an analytic approach linking the density of complex resonance poles to the distribution of reflection coefficients at complex energies, yielding explicit formulas for the crossover from narrow to broad resonances in both semi-infinite and short one-dimensional disordered samples, and validating these results against numerical simulations of the Anderson tight-binding model.

The Gist

Here is an explanation of the paper, translated from complex physics jargon into everyday language using analogies.

The Big Picture: Echoes in a Noisy Room

Imagine you are standing in a long, narrow hallway. The walls are covered in random bumps, pebbles, and uneven patches (this is the disordered medium). You shout a sound wave down the hall.

In a perfect hallway, the sound would travel straight to the end and bounce back cleanly. But in this messy hallway, the sound bounces off the bumps in a million different directions. Some of it gets trapped, some of it gets lost, and some of it eventually makes it back to you as an echo.

This paper is about understanding the nature of those echoes. Specifically, the authors want to know:

1. How long do the echoes last? (Do they fade away instantly, or do they linger for a long time?)
2. How many different "types" of echoes are there?

In physics, these lingering echoes are called resonances. The "width" of a resonance tells you how quickly the energy leaks out. A narrow width means the echo lasts a long time; a broad width means it dies out quickly.

The Problem: A Messy Math Puzzle

For decades, physicists have been great at predicting what happens when the hallway is very long and the bumps are very strong. In that case, the sound gets stuck (localized) and the echoes are very specific and long-lived.

However, nobody had a good mathematical recipe for two other situations:

1. The "Short Hallway": When the hallway is short, the sound doesn't have time to get stuck. It bounces around a few times and leaves. The math for this was missing.
2. The "In-Between": When the echoes are neither super long nor super short, but somewhere in the middle.

The authors of this paper, Yan Fyodorov and Jan Meibohm, developed a new "magic lens" to look at this problem.

The Magic Lens: Turning Sound into a Map

The authors realized they didn't need to track every single bounce of the sound wave. Instead, they found a clever shortcut.

Imagine you have a special microphone that can listen to the echo not just at your current pitch, but at slightly "absorbed" pitches (like listening to the echo while wearing noise-canceling headphones that dampen the sound).

They discovered a direct link: If you know how the reflection changes when you add a little bit of "absorption" (dampening), you can mathematically calculate exactly how many echoes of different lengths exist.

Think of it like this:

• The Old Way: Trying to count every single fish in a dark ocean by swimming around and looking for them. (Hard, slow, and you miss a lot).
• The New Way: Using sonar. You send out a ping, and the pattern of the returning sound tells you exactly how many fish are there and how big they are, without ever seeing them.

This "sonar" is their new formula. It connects the reflection coefficient (how much sound bounces back) to the density of resonances (how many echoes exist).

The Two New Discoveries

Using this new formula, they solved two major mysteries:

1. The Short Hallway (The Ballistic Regime)

When the hallway is very short (shorter than the distance a sound wave usually travels before getting confused), the sound behaves like a billiard ball. It zips back and forth a few times and leaves.

• The Finding: The authors found a new formula for this. They discovered that in short hallways, the echoes are generally "broad" (they die out fast).
• The Analogy: It's like shouting in a small bathroom. The echo is loud but disappears almost instantly. The math they derived describes exactly how this "fast fade" happens, a regime that had been ignored until now.

2. The Long Hallway (The Localized Regime)

When the hallway is very long and messy, the sound gets trapped in pockets.

• The Finding: They confirmed that most echoes are very long-lived (narrow width), but there is a "tail" of very short-lived echoes too.
• The Analogy: Imagine a maze. Most people get lost and wander for a long time (long echoes). But a few people find the exit immediately (short echoes). The authors mapped out the ratio of "lost wanderers" to "quick exiters" perfectly.

The "Crossover": The Smooth Transition

The most beautiful part of their work is that their formula doesn't just work for "short" or "long." It works for everything in between.

Imagine a dimmer switch for a light.

• Turn it all the way down: You get the "Short Hallway" physics.
• Turn it all the way up: You get the "Long Hallway" physics.
• The Paper's Contribution: They showed you exactly how the light changes as you slide the switch from one end to the other. They provided a single, unified map that describes the transition from a short, fast-fading echo to a long, lingering one.

Why Does This Matter?

You might ask, "Who cares about echoes in a messy hallway?"

Actually, this applies to almost everything that moves as a wave:

• Electrons in computer chips: If you are building a tiny computer chip, electrons act like waves. If the material is messy, the electrons get stuck, and the chip stops working. Understanding these "echoes" helps engineers design better materials.
• Light in fiber optics: Signals in the internet travel as light. If the glass fiber is imperfect, the signal degrades.
• Sound in concert halls: Architects use these principles to design rooms where sound doesn't get trapped or die out too quickly.

Summary

In simple terms, Fyodorov and Meibohm invented a new mathematical "sonar" that allows scientists to predict exactly how waves behave in messy environments. They solved the math for short environments (which was previously a mystery) and unified the math for short and long environments into one smooth, elegant formula.

They didn't just look at the problem; they built a bridge that connects the world of "fast, bouncing waves" to the world of "slow, trapped waves," showing us that they are just two sides of the same coin.

Technical Summary

Here is a detailed technical summary of the paper "Density of reflection resonances in one-dimensional disordered Schrödinger operators" by Yan V. Fyodorov and Jan Meibohm.

1. Problem Statement

The paper addresses the statistical properties of scattering resonances in one-dimensional (1D) disordered quantum systems, specifically focusing on the pure reflection problem.

• Context: In a 1D disordered medium (modeled by a Schrödinger operator with white-noise potential), wave propagation is governed by Anderson localization. While the spectral statistics of closed systems (discrete eigenvalues) are well-studied, the statistics of resonances in open systems (poles of the scattering matrix in the complex energy plane) are less understood, particularly regarding the full distribution of resonance widths.
• Objective: To develop an analytic approach to evaluate the mean density of complex resonance poles, $\rho(E, \Gamma)$, where $E$ is the real energy and $\Gamma$ is the resonance width (imaginary part of the pole). The authors aim to cover the crossover from narrow to broad resonances and address regimes previously inaccessible, such as short disordered samples (ballistic regime).

2. Methodology

The authors employ a multi-faceted approach combining analytic derivations, asymptotic analysis, and numerical simulations.

A. Fundamental Link: Resonance Density and Reflection Coefficient

The core theoretical innovation is establishing a rigorous link between the density of resonance poles and the distribution of the reflection coefficient at complex energies.

• Complex Energy: The authors introduce a complex energy $E = E + i\eta$, where $\eta > 0$ represents a uniform absorption rate in the medium.
• Main Relation: They derive the central formula (Eq. 25):
  $$\rho(E, \Gamma) = \frac{1}{4\pi} \frac{\partial^2}{\partial \eta^2} \langle \ln r(E + i\eta, L) \rangle \Big|_{\eta=\Gamma}$$
  where $r = |R(E, L)|^2$ is the reflection coefficient, and the average $\langle \dots \rangle$ is taken over disorder realizations. This reduces the problem of finding resonance poles to calculating the statistics of the reflection coefficient in the presence of absorption.

B. Stochastic Dynamics (Invariant Imbedding)

To evaluate the average $\langle \ln r \rangle$, the authors utilize the invariant imbedding method for the continuous 1D model with white-noise potential.

• Stochastic Riccati Equation: The reflection amplitude $R_k(x)$ satisfies a stochastic differential equation (SDE) driven by Brownian motion.
• Fokker-Planck Equation: By separating the phase and modulus, they derive a Fokker-Planck equation for the probability density $P(r, x)$ of the reflection coefficient $r$ as a function of sample length $x$.
• Length Scales: The analysis is governed by three key scales:
  1. Sample length $L$.
  2. Localization length $\ell_L = 4k^2/D$.
  3. Absorption length $\ell_A = k/\eta$.

C. Asymptotic Regimes

The authors solve the Fokker-Planck equation in two distinct limits:

1. Semi-infinite/Long Sample Limit ($L \gg \ell_L$):
   • They utilize the stationary solution of the Fokker-Planck equation.
   • This yields an explicit formula for $\rho(E, \Gamma)$ that describes the crossover from narrow to broad resonances.
2. Short Sample Limit ($L \ll \ell_L$, Ballistic Regime):
   • This regime had not been systematically addressed in literature.
   • The authors employ a WKB-like asymptotic expansion (in the small parameter $l = L/\ell_L$) for the Fokker-Planck equation.
   • They further develop a Laplace-transformed WKB method to solve the resulting Hamilton-Jacobi equations and perform asymptotic matching near the boundary $r=0$ (where the WKB approximation fails). This leads to an explicit formula for the resonance density in the ballistic regime.

D. Numerical Validation

To verify the analytic results, the authors perform extensive numerical simulations on the discrete Anderson tight-binding model:

• Exact Resonances: Finding roots of the determinant of the effective non-Hermitian Hamiltonian.
• Parametric Resonances: Introducing an improved approximation (Eq. 96) that retains first-order energy dependence, significantly outperforming standard parametric resonance approximations (Eq. 95).
• Direct Simulation of SDE: Solving the Langevin equation for the reflection coefficient to compute $\langle \ln r \rangle$ and applying the main relation (Eq. 25) via numerical differentiation.

3. Key Results

A. Long Sample Limit ($L \to \infty$)

The authors derive an explicit formula for the resonance density (Eq. 81):
$$\rho^{(\infty)}_k(\Gamma) = \frac{\ell_L^2}{k L} \left[ \frac{k}{\ell_L \Gamma} - 2 e^{2\ell_L \Gamma / k} E_1\left( \frac{2\ell_L \Gamma}{k} \right) \right]$$

• Narrow Resonances ($\Gamma \ll D/k$): The density scales as $\rho \sim 1/\Gamma$. This is consistent with the exponential localization of eigenfunctions.
• Broad Resonances ($\Gamma \gg D/k$): The density decays as $\rho \sim 1/\Gamma^2$.
• Cutoff: A natural cutoff scale $\Gamma_{\min} \sim \Gamma_L e^{-L/\ell_L}$ is identified, below which the density is suppressed due to rare events (eigenstates localized far from the open boundary).

B. Short Sample Limit ($L \ll \ell_L$)

For short samples, the authors derive a new limiting formula (Eq. 87).

• Typical Width: Resonances are broad, with typical widths scaling as $\Gamma_S = k/L$.
• Behavior: The density exhibits a sharp peak around $\Gamma \sim \Gamma_S$ and follows a universal $1/\Gamma^2$ tail for very broad resonances.
• Novelty: This regime was previously unanalyzed; the derivation required non-trivial WKB asymptotics and boundary layer matching.

C. Numerical Findings

• The improved parametric resonance approximation (Eq. 96) is shown to be significantly more accurate than standard methods, especially in the ballistic regime.
• Numerical results for both short and long samples show excellent agreement with the derived analytic formulas (Eqs. 81 and 87).
• The method of computing $\rho$ via numerical differentiation of $\langle \ln r \rangle$ (using the main relation) is validated as a viable alternative to direct pole finding.

4. Significance and Contributions

1. Unified Analytic Framework: The paper provides a unified analytic description of resonance densities across the entire spectrum of disorder strengths and sample sizes, bridging the gap between narrow (localized) and broad (ballistic) resonances.
2. New Regime Analysis: It is the first systematic study of resonance statistics in the short-sample (ballistic) limit, revealing a distinct scaling behavior and a sharp peak in the width distribution.
3. Methodological Advancement: The derivation of the link between resonance density and the second derivative of the log-reflection coefficient (Eq. 25) offers a powerful tool for future studies, potentially applicable to higher dimensions or different disorder types.
4. Improved Numerical Schemes: The introduction of the "improved parametric resonance" approximation provides a computationally efficient and highly accurate method for estimating resonance statistics in discrete models, surpassing existing techniques.
5. Physical Insight: The results clarify the relationship between Anderson localization, absorption, and the temporal decay of scattered signals (via resonance widths), offering deeper insight into transport in disordered media.

In summary, the paper successfully combines rigorous mathematical physics (Fokker-Planck equations, WKB, complex analysis) with numerical verification to solve a long-standing problem in the statistical theory of disordered systems, providing explicit formulas for resonance densities in regimes previously inaccessible to analytic treatment.