Unidirectional Entropic Solutions of the Pressureless Euler Alignment System

Technical Summary: Unidirectional Entropic Solutions of the Pressureless Euler Alignment System

Problem Statement
The paper investigates the pressureless Euler Alignment system, a continuum model describing the collective behavior of agents interacting through a nonlocal communication protocol. The system is given by:
$$\begin{cases} \partial_t\rho + \nabla\cdot (\rho u) = 0 \\ \partial_t(\rho u) + \nabla\cdot (\rho u \otimes u) = \int_{\mathbb{R}^d} \phi(x - \xi)(u(\xi) - u(x))\rho(x)\rho(\xi) d\xi \end{cases}$$
While well-posedness is well-understood in one dimension ($d=1$) and for specific 1D reductions, the general multidimensional case remains largely open, particularly regarding the continuation of solutions past blow-up times (loss of regularity). The authors focus specifically on unidirectional dynamics, where the velocity field is constrained to a single direction, $u(x, t) = (u(x, t), 0, \dots, 0)$. In this setting, particle trajectories preserve ordering along the flow direction but are coupled nonlocally across transverse directions. The primary challenge is establishing the existence, uniqueness, and stability of weak (entropy) solutions in this multidimensional, unidirectional regime under bounded, Lipschitz communication protocols.

Methodology
The authors employ a multi-step analytical framework that bridges discrete particle dynamics and continuum weak solutions:

1. Scalar Balance Law Formulation:
   The system is recast as a family of coupled scalar balance laws parametrized by the transverse variable $y \in \mathbb{R}^{d-1}$. By disintegrating the density measure $\rho$ with respect to horizontal slices and defining a cumulative distribution function $M(x, y, t)$ and a flux function $A(M, y)$, the Euler Alignment system reduces to:
   $$\partial_t M + \partial_x (A(M, y)) = B[\nu, M] \partial_x M$$
   where $B[\nu, M]$ is a nonlocal term representing the alignment force, and $\nu$ is the transverse measure (the projection of $\rho$ onto $\mathbb{R}^{d-1}$).

2. Entropy Conditions:
   The authors define entropy solutions in the sense of Kruzkov, adapted to the nonlocal setting. The entropy inequality involves a source term arising from the nonlocal coupling $B[\nu, M]$.

3. Stability Estimates via Doubling of Variables:
   To prove uniqueness and stability, the authors utilize the Kruzkov doubling-of-variables technique. A critical technical innovation is the formulation of stability estimates involving the optimal coupling $\pi$ between two distinct transverse measures $\nu$ and $\tilde{\nu}$. They derive two complementary stability estimates:

   • Linear-in-time estimate: Depends on the Lipschitz seminorm of the flux difference. This is insufficient for their existence proof because the approximating fluxes do not necessarily converge in the Lipschitz seminorm.
   • Exponential-in-time estimate: Relies only on the $L^\infty$ difference of the fluxes (inspired by Bouchut and Perthame). This low-regularity estimate is essential for passing to the limit in their approximation scheme.

4. Approximation Scheme:
   Existence is constructed via a two-level discretization:

   • Transverse Discretization: The continuous transverse measure $\nu$ is approximated by a sequence of atomic measures $\nu_k$ (discretizing the $y$-direction).
   • Longitudinal Discretization: For each atomic slice, the solution is approximated by Sticky Particle Cucker–Smale (SPCS) dynamics, where agents follow ODEs and stick together upon collision (inelastic collisions conserving momentum).
     The authors prove that the sequence of SPCS solutions converges to the unique entropy solution of the scalar balance law, which in turn generates the weak solution to the Euler Alignment system.

Key Contributions and Results

• Well-posedness for Unidirectional Solutions: The paper establishes the existence, uniqueness, and stability of entropy solutions for the unidirectional Euler Alignment system in $\mathbb{R}^d$ ($d \geq 2$) under the assumption of a bounded, Lipschitz communication protocol. This extends the 1D theory to a setting that retains the ordering property of 1D but introduces rich nonlocal coupling in transverse directions.
• Novel Stability Framework: The authors introduce a stability framework based on the Wasserstein-1 distance between transverse measures and optimal couplings. This allows for the comparison of solutions with different transverse supports, a necessary step for the approximation by atomic measures.
• Convergence Rates: Under additional Hölder regularity assumptions on the initial data and flux, the authors provide explicit rates of convergence for the approximation of general weak solutions by sticky particle dynamics.
• Flocking with Degenerate Protocols: The paper investigates long-time behavior (flocking). A significant finding is that flocking (velocity alignment and bounded spatial diameter) occurs even if the communication protocol vanishes in a cylindrical neighborhood of the flow axis, provided the protocol is "heavy-tailed" outside this cylinder and the initial mass distribution is not concentrated within two such cylinders. This demonstrates that direct communication along the direction of motion is not strictly necessary for flocking in unidirectional systems.

Significance and Claims
The authors claim that this work provides the first comprehensive theory for continuing unidirectional Euler Alignment solutions past the blow-up time in dimensions $d \geq 2$. The significance lies in:

1. Bridging Dimensions: It treats the unidirectional case as a "family of coupled 1D problems," capturing the richness of higher-dimensional emergent structures while leveraging the technical advantages of 1D ordering.
2. Robustness of Flocking: The flocking result challenges the intuition that alignment requires communication along the direction of motion, showing that transverse communication suffices if the geometry allows.
3. Methodological Advancement: The development of stability estimates that rely on $L^\infty$ flux control rather than Lipschitz control is presented as a crucial technical tool that enables the construction of solutions via discrete approximations where flux regularity cannot be guaranteed.

The paper remains modest regarding the scope of its assumptions, noting that the results currently require bounded, Lipschitz protocols and do not yet cover strongly singular protocols (though they speculate on future extensions). The work is framed as a foundational step toward understanding weak solutions in multidimensional collective dynamics.