This paper presents a QMU-native model for ball lightning as a stable, spherical luminous plasmoid governed by a unit-level universal invariant. Working strictly in Quantum Measurement Units (QMU), the core result is a factorization of the ball-lightning state descriptor: = e₄₌₀ₗ^2\, C^2\, Fq^3= (e₄₌₀ₗ^2₂) \, (mflx curl rson), yields the falsifiable onset invariant\ball{mflx curl rson = e₄₌₀ₗ^2₂} to be constant (within uncertainty) across gas species, pressure, and chamber geometry at the formation threshold of a stable spherical plasmoid. A minimal quadratic Lagrangian is constructed in QMU for a single-mode quasi-spherical configuration, leading to equilibrium and small-oscillation stability conditions and an operational threshold of the form curl rson C₁₋, (₄₌₀ₗ^2₂), a dimensionless constant C₁₋ to be measured (canonically approaching unity for the symmetric spherical solution). Extensions are provided for an explicit chronovibrational degree of freedom and for multi-mode coupling; these modify the measured threshold constant but preserve the factorized structure. The paper is designed for laboratory falsifiability. It provides step-by-step protocols for (i) plasma onset mapping in the (mflx, curl, freq) space, (ii) ball formation and invariant clustering tests at onset, and (iii) thermoradiative post-onset diagnostics using the QMU temperature channel. Recommended data products, closure diagnostics, and a representative uncertainty budget are included so independent groups can reproduce and audit the invariant test. If validated, the invariant elevates “ball lightning” from a descriptive label to a metrologically closed state descriptor in QMU: a spherical-charge seed factor (e₄₌₀ₗ^2₂) multiplied by the dynamical product (mflx) that can be directly targeted and measured in controlled experiments.
Thomson et al. (2025) studied this question.