Beyond the pointwise classification of SU(2) gauge fields: local invariant completeness, differential assembly, dynamical realization, and the degenerate orbit structure of the Wang–Yang classes

At a single space-time point an SU(2) gauge-field strength is encoded by the complex 3×3 matrix $A=E+iH$ built from the electric-type and magnetic-type isovectors; Wang and Yang (1978) completed its classification under local gauge rotations and proper orthochronous Lorentz transformations through the rank of the quadratic invariant $\Delta=\tilde{A}A$: full-rank $\Delta$ admits exactly two gauge-and-Lorentz-inequivalent realizations, rank 2 a unique one, while the degenerate rank-1 and rank-0 strata carry continuous families of inequivalent realizations (in subcase 3a a complex family parametrized by the ratio $\lambda:\mu$, in subcase 3b a real $\mu$-family with the identification $\mu\sim 1/\mu$, and in the null case a real $\mu$-family with the identification $\mu\mapsto -1/\mu$). The immediate follow-ups of 1977–1984 — Roskies's polynomial invariant basis, Carmeli's eigenspinor–eigenvalue scheme, the Anandan–Tod spinor classification, the little-group classification of Castillejo–Kugler–Roskies, Cho's canonical forms, and Urbantke's infinitesimal differential classification — refined the pointwise theory while explicitly deferring four interlinked questions: how pointwise classes assemble on space-time regions under derivative compatibility conditions; which classes and continuous parameter values are realized by solutions of the source-free Yang–Mills equations; whether local gauge-and-Lorentz invariant data beyond $\Delta$ can separate the continuous degenerate families; and the complete orbit and identification structure on the degenerate strata. Modern work that touches these questions (Nayak–Shrock 2008; Dietrich 2009; Garat 2013–2025) remains at the pointwise level, and the sectoral classifications of all constant solutions (Shirokov 2023) and of gauge fields with vanishing scalar invariants (Kuchynka 2019) cover only special sectors of the dynamical and invariant questions. A precise, nonempty open core survives.

Background

An SU(2) Yang–Mills gauge field is a connection on a principal SU(2) bundle over space-time; its field strength is a Lie-algebra-valued antisymmetric tensor $F^{a}_{\mu\nu}$ (the internal index $a=1,2,3$ labels isospin, and $\mu,\nu=0,1,2,3$ label space-time) satisfying the source-free field equation $D_\mu F^{\mu\nu}=0$ and the non-Abelian Bianchi identity $D_{[\mu}F_{\nu\rho]}=0$, where $D$ is the gauge-covariant derivative. At one space-time point the field strength can be encoded in two triples of ordinary 3-vectors carrying an internal index: the electric-type isovector $E^{a}_{i}=F^{a}_{0i}$ and the magnetic-type isovector $H^{a}_{i}=\tfrac{1}{2}\epsilon_{ijk}F^{a}_{jk}$. Wang and Yang combine them into the 3×3 complex matrix $A=E+iH$, a non-Abelian analogue of the Riemann–Silberstein vector of electromagnetism (the complex combination $\mathbf{E}+i\mathbf{H}$ of the electric and magnetic fields, which packages the two fields so that Lorentz transformations act on it by complex rotations). A proper orthochronous Lorentz transformation (one of determinant 1 that preserves the direction of time) acts on $A$ by left multiplication with a complex orthogonal matrix $L$ ($\tilde{L}L=1$, $\det L=1$, tilde denoting transpose) — a correspondence proved in an appendix of the paper — and a local gauge transformation at the point acts by right multiplication with a real orthogonal $G\in SO(3)$ arising from the adjoint action of SU(2) on its own Lie algebra through the double covering $SU(2)\to SO(3)$. The pointwise classification problem is the orbit problem for this combined group action.

The only quadratic Lorentz invariant is the complex symmetric matrix $\Delta=\tilde{A}A$; under gauge rotations it transforms by congruence $\Delta\mapsto\tilde{G}\Delta G$, and its gauge-independent content comprises 9 real parameters. The electromagnetic analogue makes the strategy plain: there the single complex invariant $\Delta=K+iJ$, with $K=|E|^{2}-|H|^{2}$ and $J=2E\cdot H$, determines the field uniquely up to Lorentz transformations when $\Delta\neq 0$, while $\Delta=0$ yields only the vacuum and a single radiation-like class. Wang and Yang proved the SU(2) counterpart: if $\operatorname{rank}\Delta=3$, exactly two inequivalent upper-triangular standard forms realize $\Delta$ (differing by the sign of one diagonal entry); if $\operatorname{rank}\Delta=2$, a unique standard form; if $\operatorname{rank}\Delta=1$ — in which case $\Delta$ can be reduced by a gauge rotation to the outer-product form $\Delta=v^{\mathrm T}v$ of a complex row vector $v=(a,f,0)$ with $a\neq 0$ fixed up to sign by a phase convention of the paper — realizations fall into subcase 3a (ratio $f/a$ non-real), a complex one-parameter family parametrized by the ratio $\lambda:\mu$ (only the ratio is meaningful), and subcase 3b ($f/a$ real), a real one-parameter family with $\mu\in[-1,1]$ and the identification $\mu\sim 1/\mu$; and if $\Delta=0$ (the null or radiation-like stratum), either the vacuum or a one-parameter family with $\mu\in[0,1]$ and the identification $\mu\mapsto -1/\mu$. The degenerate strata are the crux: there the quadratic invariant fails to characterize the field, and continuous families of inequivalent local field strengths survive — a phenomenon absent in electromagnetism.

The paper sits in a cluster of contemporaneous classifications of the same object: Carmeli's eigenspinor–eigenvalue scheme (a spinor-based classification of the field strength); the spinor and rank-based classification of Anandan and Tod for arbitrary gauge groups; the little-group classification of Castillejo, Kugler, and Roskies, which labels a field by its little group — the subgroup of the combined Lorentz and gauge group that leaves the field invariant; Cho's canonical forms and frames for arbitrary groups; and Anandan's multiplicity analyses. On the invariant side, Roskies and Anandan–Roskies constructed a polynomial basis for the local Yang–Mills invariants: every polynomial invariant of the SU(2) field strength at a point is a polynomial in ten basic invariants, of which nine are functionally independent. On the differential side, Urbantke began extending the pointwise view: he studied the 2-plane elements on which infinitesimal parallel displacement of isovectors yields the identity, obtaining an algebraic and differential classification and, in the generic case, a conformal quasimetric naturally associated with the field; whether such a plane distribution is Frobenius-integrable (i.e., tangent to a family of two-dimensional surfaces on which parallel transport is path-independent) is exactly the kind of compatibility condition that a pointwise classification ignores. Asymptotic peeling theorems (Roskies; Anandan–Tod) — the gauge-field analogue of the peeling of the Weyl curvature in general relativity, describing which components of the field dominate the fall-off along outgoing null rays at null infinity — further constrain which classes dominate the field in the radiative limit.

What remains unresolved is the extension of this classification beyond the pointwise quadratic level: the assembly of the pointwise classes into fields on regions (Bianchi-identity and potential-representability compatibility), the dynamical selection of the continuous parameters by the field equations, the completeness of local invariants on the degenerate strata, and the full orbit and identification structure there. These are the open directions.

Problem Statement

Extend the completed pointwise classification of SU(2) gauge-field strengths — the orbit structure of the complex matrix $A=E+iH$ under local gauge rotations and proper orthochronous Lorentz transformations, organized by the rank of the quadratic invariant $\Delta=\tilde{A}A$ — to a complete local and differential theory of the field, by resolving the following interlinked directions.

(a) Differential assembly: characterize which assignments of pointwise classes (rank of $\Delta$, normal-form parameters $\lambda$, $\mu$) to the points of a space-time region are realized by an actual SU(2) field strength, i.e. by the curvature $F=\mathrm{d}\mathcal{A}+\mathcal{A}\wedge\mathcal{A}$ of some gauge potential $\mathcal{A}$ (a symbol distinct from the matrix $A=E+iH$ of the pointwise theory), satisfying the non-Abelian Bianchi identity; that is, give the compatibility and integrability theory for smoothly assembling the pointwise normal forms on regions, including the Frobenius-type integrability of the 2-plane distributions that Urbantke associated with the field and the obstructions arising on the degenerate strata.

(b) Dynamical realization: determine which pointwise classes — in particular which values of the continuous parameters (the ratio $\lambda:\mu$ in subcase 3a, and $\mu$ in subcase 3b and in the rank-0 null family) — are realized at points of solutions of the source-free SU(2) Yang–Mills equations $D_\mu F^{\mu\nu}=0$, distinguishing purely local realizability from realizability under physically relevant global conditions such as finite energy or asymptotic radiation falloff, and identify the dynamical or geometrical criteria that select them.

(c) Invariant completeness: determine whether local gauge-and-Lorentz-invariant data beyond the quadratic $\Delta$ — higher-order polynomial invariants (such as the ten basic invariants of Anandan–Roskies) or local data involving derivatives of the field — can separate the continuous families of inequivalent realizations that share a degenerate $\Delta$, or prove that no such local data can; equivalently, characterize how much of the gauge-and-Lorentz orbit structure is measurable by local invariants.

(d) Orbit structure on the degenerate strata: complete the pointwise orbit structure that the Wang–Yang classification left open — give a complete group-theoretic description of the orbit space of the field-strength space under the combined local gauge $SO(3)$ and Lorentz action on the rank-1 and null strata, and determine whether the known discrete parameter identifications ($\mu\sim 1/\mu$ in subcase 3b, $\mu\mapsto -1/\mu$ in the rank-0 case) are exhaustive, i.e. whether any further discrete identifications occur.

The enumerated clauses within a direction are jointly posed parts of that single question.

References

1. Ling-Lie Wang and Chen Ning Yang, "Classification of SU(2) gauge fields", Physical Review D 17, 2687 (1978). DOI: 10.1103/PhysRevD.17.2687, https://doi.org/10.1103/PhysRevD.17.2687
2. Ralph Roskies, "Invariants and Classification of Yang-Mills Fields", Physical Review D 15, 1722 (1977). DOI: 10.1103/PhysRevD.15.1722, https://doi.org/10.1103/PhysRevD.15.1722
3. Moshe Carmeli, "Classification of Yang-Mills Fields", Physical Review Letters 39, 523 (1977). DOI: 10.1103/PhysRevLett.39.523, https://doi.org/10.1103/PhysRevLett.39.523
4. J. Anandan and K. P. Tod, "Invariants and Classification of Gauge Fields", Physical Review D 18, 1144 (1978). DOI: 10.1103/PhysRevD.18.1144, https://doi.org/10.1103/PhysRevD.18.1144
5. J. Anandan and R. Roskies, "Polynomial Basis for Yang-Mills Invariants", Journal of Mathematical Physics 19, 2614 (1978). DOI: 10.1063/1.523616, https://doi.org/10.1063/1.523616
6. J. Anandan and R. Roskies, "Spinorial Classification of the SU(2) Gauge Field", Physical Review D 18, 1152 (1978). DOI: 10.1103/PhysRevD.18.1152, https://doi.org/10.1103/PhysRevD.18.1152
7. M. Carmeli, "Classification of gauge fields: The eigenspinor-eigenvalue equation", Physics Letters B 77, 188 (1978). DOI: 10.1016/0370-2693(78)90618-4, https://doi.org/10.1016/0370-2693(78)90618-4
8. M. Carmeli and M. Fischler, "Classification of SU(2) gauge fields: Lorentz invariant versus gauge invariant schemes", Physical Review D 19, 3653 (1979). DOI: 10.1103/PhysRevD.19.3653, https://doi.org/10.1103/PhysRevD.19.3653
9. L. Castillejo, M. Kugler, and R. Z. Roskies, "A Little Group Classification of Gauge Fields", Physical Review D 19, 1782 (1979). DOI: 10.1103/PhysRevD.19.1782, https://doi.org/10.1103/PhysRevD.19.1782
10. Y. M. Cho, "Classification of gauge fields", Journal of Mathematical Physics 20, 2605 (1979). DOI: 10.1063/1.524022, https://doi.org/10.1063/1.524022
11. J. Anandan, "Classification of gauge fields", Journal of Mathematical Physics 20, 260 (1979). DOI: 10.1063/1.524075, https://doi.org/10.1063/1.524075
12. M. Carmeli and D. H. Wohl, "Chern classes, Pontrjagin forms and classification of gauge fields", Lettere al Nuovo Cimento 25, 230 (1979). DOI: 10.1007/BF02776232, https://doi.org/10.1007/BF02776232
13. M. Carmeli and B. Z. Moroz, "Classification of gauge fields and group representations", Lecture Notes in Mathematics 836, 313 (1980). DOI: 10.1007/BFB0089746, https://doi.org/10.1007/BFB0089746
14. M. Carmeli and Kh. Huleihil, "Type-D solution of sourceless Yang-Mills equations", Il Nuovo Cimento A 74, 245 (1983). DOI: 10.1007/BF02770878, https://doi.org/10.1007/BF02770878
15. H. Urbantke, "On integrability properties of SU(2) Yang-Mills fields. I. Infinitesimal part", Journal of Mathematical Physics 25, 2321 (1984). DOI: 10.1063/1.526402, https://doi.org/10.1063/1.526402
16. Gouranga C. Nayak and Robert Shrock, "Gauge-invariant quantities characterizing gauge fields in chromodynamics", Physical Review D 77, 045008 (2008). DOI: 10.1103/PhysRevD.77.045008, arXiv:0711.2759, https://arxiv.org/abs/0711.2759
17. Dennis D. Dietrich, "On invariants for particle propagation in non-Abelian fields", Physical Review D 80, 067701 (2009). DOI: 10.1103/PhysRevD.80.067701, arXiv:0903.2293, https://arxiv.org/abs/0903.2293
18. Alcides Garat, "Gauge invariant method for maximum simplification of the field strength in non-Abelian Yang-Mills theories", International Journal of Geometric Methods in Modern Physics 12, 1550104 (2015). DOI: 10.1142/S0219887815501042, arXiv:1306.2174, https://arxiv.org/abs/1306.2174
19. Gouranga C. Nayak, "General form of the color potential produced by color charges of the quark", Journal of High Energy Physics 2013, 001 (2013). DOI: 10.1007/JHEP03(2013)001, arXiv:1201.2666, https://arxiv.org/abs/1201.2666
20. Martin Kuchynka, "Gauge fields with vanishing scalar invariants", Classical and Quantum Gravity 36, 205016 (2019). DOI: 10.1088/1361-6382/ab4360, arXiv:1902.07281, https://arxiv.org/abs/1902.07281
21. Dmitry Shirokov, "Classification of all constant solutions of SU(2) Yang–Mills equations with arbitrary current in pseudo-Euclidean space $\mathbb{R}^{p,q}$", Modern Physics Letters A 38, 2350096 (2023). DOI: 10.1142/S0217732323500967, arXiv:1912.04996, https://arxiv.org/abs/1912.04996
22. G. Rudolph, M. Schmidt, and I. P. Volobuev, "Classification of Gauge Orbit Types for SU(n)-Gauge Theories", Mathematical Physics, Analysis and Geometry 5, 201 (2002). DOI: 10.1023/A:1020968206969, https://doi.org/10.1023/A:1020968206969