🔢 mathematics

A simple derivation of Anderson-Hartree equations with harmonic potential

This paper demonstrates that the mean-field limit and the zero-noise limit can be commuted for Anderson-Hermite operators in dimensions 1, 2, and 3 by adapting the Knowles-Pickl derivation and establishing uniform energy estimates for the resulting Anderson-Hartree equations.

Original authors: Samaël Mackowiak

Published 2026-08-25

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Samaël Mackowiak

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

In the vast, quiet realm of quantum physics, scientists often face a dilemma of scale. On one side, there is the microscopic world of individual particles, where the rules are governed by complex, linear equations that describe how a single electron or atom moves. On the other side, there is the macroscopic world of matter as we see it, where billions of particles interact, collide, and influence one another. Bridging these two worlds is one of the most enduring challenges in physics: how do we derive the simple, collective behavior of a large group from the chaotic, individual dance of its members? For decades, researchers have developed mathematical tools to show that when a huge number of particles interact weakly, their collective motion can be described by a single, simpler equation. This process, known as the mean-field limit, allows physicists to replace a tangled web of individual interactions with a smooth, average field that every particle feels. However, real-world environments are rarely perfect. Particles often move through disordered materials, encountering random impurities or fluctuations that act like a chaotic, invisible fog. This randomness, known as noise, makes the mathematical description of the system significantly harder, as the usual tools for simplifying the equations often break down when the environment is this unpredictable.

A recent study by S. Mackowiak addresses this specific challenge by showing how to mathematically derive a simplified equation for a large group of particles moving through such a noisy, disordered environment. The researcher focused on a system where particles are trapped in a potential well, a sort of invisible bowl that keeps them confined, while simultaneously being buffeted by a random, fluctuating force. The goal was to prove that even with this randomness, the collective behavior of a massive number of these particles can still be accurately described by a single, effective equation, provided the number of particles is large enough. The study successfully demonstrates that the order of two different mathematical limits can be swapped: one can first let the number of particles grow to infinity to find the average behavior, and then let the randomness settle, or vice versa, and arrive at the same result. This is a crucial step because it confirms that the simplified model is robust and reliable, even when the underlying environment is messy and uncertain.

The work begins with a system of many particles, each described by a wave that evolves over time. In a perfect world without disorder, the interaction between these particles is often modeled as a smooth, average force. However, when the particles move through a medium with random imperfections, the equations become much more difficult to solve. The researcher had to deal with two distinct mathematical hurdles at once. The first was the sheer number of particles, which requires a transition from a many-body description to a single-particle description. The second was the nature of the noise itself, which is so irregular that it cannot be treated as a standard function; it requires a special kind of mathematical handling to make sense of it. The study focuses on dimensions one, two, and three, covering the physical spaces we inhabit. In the lower dimensions, the mathematical objects representing the noise can be constructed rigorously. In three dimensions, the situation is more complex, and the researcher treats the noise mathematically as a formal object, acknowledging that a full, rigorous construction of the three-dimensional version is a task for future work.

To solve this, the researcher adapted a known method for deriving these simplified equations, originally developed for clean systems, and modified it to work with this noisy, disordered setup. The core of the proof involved showing that the energy of the system remains under control, even as the noise fluctuates. By proving that the solutions to the complex, many-particle equations stay within certain bounds regardless of the specific realization of the noise, the researcher could demonstrate that the system behaves predictably in the long run. This allowed for a "double limit" to be taken: first, the number of particles was sent to infinity, and second, the approximation of the noise was refined to its true, random form. The result is a new, simplified equation that describes the evolution of the entire cloud of particles as a single entity, influenced by both the confining trap and the random noise.

The findings confirm that this simplified equation, which includes a term for the random noise and a term for the average interaction between particles, is the correct description of the system's behavior. The study proves that the difference between the actual many-particle system and this simplified model shrinks to zero as the number of particles increases, and this convergence happens uniformly, meaning it holds true for all reasonable variations of the noise and the interaction strength. This is a significant achievement because it validates the use of these simplified models in real-world scenarios where disorder is present, such as in certain types of condensed matter physics or quantum optics. The researcher also showed that the mathematical space in which these solutions exist is well-behaved, allowing for the existence and uniqueness of the solution to the simplified equation. This means that for any given starting condition, there is exactly one way the system will evolve, and that evolution is stable.

One of the most practical aspects of this work is that it handles a wide variety of interaction forces between the particles, including those that mimic the electrical attraction or repulsion seen in nature, without requiring the forces to be artificially smoothed out. The study shows that the method works even for these physically relevant, long-range forces. By establishing that the mean-field limit and the limit of vanishing noise regularization can be commuted, the paper provides a solid theoretical foundation for using these effective equations to study complex quantum systems in disordered environments. It confirms that the complexity of the individual random fluctuations does not destroy the collective order that emerges in large systems. Instead, the collective behavior remains robust, governed by a single, deterministic equation that captures the essence of the noisy, many-body reality. This result offers a clear path forward for physicists who wish to model quantum systems in realistic, imperfect conditions, ensuring that their simplified models are not just convenient approximations, but mathematically sound descriptions of nature.

Technical Summary: Derivation of Anderson-Hartree Equations with Harmonic Potential

Problem Statement
This paper addresses the mathematical derivation of the Anderson-Hartree equation from the many-body Schrödinger equation in the presence of spatial white noise. Specifically, the author seeks to justify the mean-field limit for a system of $N$ interacting bosons subject to a random potential (Anderson disorder) and a harmonic confining potential.

The starting point is the $N$ -body equation (1.2):
$i\partial_t\Psi_N + \sum_{j=1}^N (H + \xi_\epsilon)_{x_j}\Psi_N + \frac{1}{N}\sum_{1\le i<j\le N} w(x_i - x_j)\Psi_N = 0$
where $H = \Delta - |x|^2$ is the harmonic oscillator Hamiltonian, $\xi$ is spatial white noise, $\xi_\epsilon$ is a smooth approximation, and $w$ is an interaction kernel. The target equation is the nonlinear Anderson-Hartree equation (1.1):
$i\partial_t u + (H + \xi)u + (w * |u|^2)u = 0$
The core challenge lies in handling a "double limit": the mean-field limit as the number of particles $N \to \infty$ and the noise regularization limit as the approximation parameter $\epsilon \to 0$ (or equivalently, the limit of the noise parameter $\theta$ ). The paper aims to prove that these limits can be commuted, establishing the convergence of the many-body marginals to the tensor product of the solution of the limit equation.

Methodology
The derivation relies on a combination of functional analysis, stochastic PDE techniques, and the mean-field convergence method developed by Pickl.

Abstract Framework for Anderson Operators:
The paper constructs the Anderson-Hermite operator $A_\theta = -(H + \xi)$ via its quadratic form. In dimensions $d=1$ and $d=2$ , the operator is rigorously defined using an exponential transform $u = e^Y v$ , where $Y$ is related to the noise $\xi$ . This transforms the singular operator into a form involving bounded coefficients and renormalized products (e.g., $Z$ ).
- In $d=1$ , the construction is direct.
- In $d=2$ , renormalization is required to define the square of the gradient of the noise potential.
- In $d=3$ , the paper treats the operator formally, assuming the existence of renormalized stochastic objects ( $Y, Z$ ) without constructing them explicitly. The operator is defined on a weighted Sobolev space $D_{1,2} = \rho W^{1,2}$ where $\rho = e^Y$ .
Well-posedness and Energy Estimates:
The author establishes global well-posedness for the limit Anderson-Hartree equation (1.3) in the space $Q(A_\theta) = D_{1,2}$ .
- Abstract Well-posedness: Using a compactness argument (Aubin-Lions-Simon lemma) and fixed-point methods, the paper proves existence and uniqueness of solutions for interaction kernels $w$ in a specific space $W$ .
- Uniform Energy Estimates: A critical step is proving that the $D_{1,2}$ -norm of the solution $u_{\theta, w}$ is uniformly bounded with respect to the noise parameter $\theta$ and the interaction $w$ on bounded sets. This is achieved by deriving energy estimates that depend on the parameters of the quadratic form (specifically the coercivity constant $c_\theta$ and the bound $C_\theta$ ), which are shown to be bounded on bounded sets of the parameter space $\Theta$ .
Mean-Field Convergence (Pickl's Method):
The convergence of the $N$ -body system to the mean-field limit is proven using a variant of Pickl's method (specifically Theorem 3.1 and Remark 3.8 from [20]).
- The method tracks the quantity $\alpha_N(t) = \|q(t)\Psi_N(t)\|_{L^2}^2$ , where $q(t)$ is the projection onto the orthogonal complement of the one-particle state $u(t)$ .
- An exponential Grönwall-type inequality is derived for $\alpha_N(t)$ , showing that $\alpha_N(t) \le C \exp(\int |u|^2) (\alpha_N(0) + 1/N)$ .
- Since the initial state is a product state, $\alpha_N(0) = 0$ . The convergence rate is thus $O(1/N)$ .

Key Contributions and Results

Theorem 1.1 (Main Result): For dimensions $d \in \{1, 2, 3\}$ , the $k$ -particle marginals $\Gamma_{k,N,\theta,w}(t)$ of the many-body solution converge to the tensor product of the limit solution $|u_{\theta,w}(t)\rangle\langle u_{\theta,w}(t)|$ in the trace-class norm. Crucially, this convergence is uniform with respect to time $t$ , the noise parameter $\theta$ , and the interaction kernel $w$ on bounded sets.
Commutativity of Limits: The uniformity of the estimates allows the author to commute the limits $N \to \infty$ and the noise regularization limit. This implies that one can pass to the limit along sequences $(\theta_N, w_N)$ converging to $(\theta, w)$ , effectively justifying the derivation of the Anderson-Hartree equation from the regularized many-body system.
Handling Dimension 3: The paper extends the method to dimension 3 by introducing a class of "formal Anderson-Hermite operators." While the stochastic construction of the 3D operator is not performed, the paper demonstrates that if such renormalized objects exist, the mean-field derivation holds.
Interaction Kernels: The work identifies a class of admissible interaction kernels $W$ (defined in Eq. 2.7) that includes physically relevant potentials. Notably, the paper highlights that the Coulomb potential $w(x) \propto |x|^{-1}$ in dimension 3 belongs to this class, allowing the treatment of Coulomb interactions without requiring Strichartz estimates.

Significance and Claims
The paper claims to provide a "simple derivation" that avoids advanced dispersive tools (like Strichartz estimates), relying instead on energy estimates and compactness arguments.

Comparison to Existing Literature: The author notes that previous work on Anderson-Hartree equations (e.g., [7]) used the BBGKY hierarchy and was limited to bounded interactions due to difficulties in proving uniqueness for the limit hierarchy. In contrast, this paper uses Pickl's method to handle more general interactions (including Coulomb in 3D) and establishes uniform bounds in the noise parameter.
Robustness: The result is robust regarding the noise parameter, proving that the mean-field limit holds uniformly for bounded sets of noise realizations.
Scope: The work is twofold: it derives the effective nonlinear equation from the linear many-body system and simultaneously provides global well-posedness results for the Anderson-Hartree equation in adapted functional spaces.

The author explicitly states that the 3D result relies on a formal construction of the Anderson-Hermite operator and that the rigorous construction of the stochastic objects in 3D is left for future work. The paper does not claim to solve the Anderson-Gross-Pitaevskii equation (cubic nonlinearity) in 3D, but rather focuses on the Hartree (convolution) nonlinearity.

More like this