Global smoothness for relativistic Vlasov–Maxwell

The explanation

In relativity, a particle's speed is u = v / √(1 + |v|²), where v is its momentum and light has speed 1Source 1. With momentum 1, the speed is about 0.707. With momentum 10, it is about 0.995. With momentum 100, it is about 0.99995. So speed never reaches 1, however hard the particle is pushed. But momentum itself can grow. The worry is that the fields push some particles harder and harder, until their momentum becomes infinite in finite time.

The relativistic Vlasov–Maxwell system models a collisionless plasma. A density f tracks how many particles sit at each position and momentum. The particles move under the electric and magnetic fields they create together, and those fields obey Maxwell's equationsSource 2. Wollman proved that smooth solutions exist for a short timeSource 3. Glassey and Strauss showed that a smooth solution continues as long as particle momenta stay boundedSource 4. Since then, global results needed small data, fewer dimensions, symmetry or a nearby known solutionSource 5.

This paper proves the general case in three dimensions. Every smooth admissible start gives a unique solution for all time. It stays smooth on every finite time intervalSource 6. Admissible means the particle density is smooth and zero outside a bounded region, and the fields are smooth with finite energy. The data may be as large as you like, with no symmetrySource 7.

The key estimate is on the signed change of a particle's momentum over time. The norm is taken only after adding up the pushes, so opposite pushes can cancelSource 8. This gives a lower bound on the time needed to double the largest momentum. These lower bounds have a sum that never ends, so momentum cannot become infinite in finite timeSource 9. As an illustration only, suppose step j took at least 1/j time units. Then 10 doublings take at least 2.93 units, and 100 take at least 5.19. The obstacle is a narrow range of angles where simple force bounds lose a factor √w at energy wSource 10. There the paper integrates the force along a source path first, and a cancellation recovers the loss.

How sure is this? Lean, a computer proof checker, has a formal statement of the main result. It covers existence, uniqueness and smoothness, with no smallness, symmetry or neutrality assumption. So the paper proves this result, as far as the Lean statement matches the paper. No specialist review is published.

What it doesn't show. This does not cover two or more species of particles, or particle data without compact support. It gives no bound on how the solution behaves as time goes to infinity; the momentum bound may grow with time.