1
$\begingroup$

In gauge theory, all measurable physical quantities remain invariant under a gauge transformation. I have always seen that the curvature 2-form $\Omega \in \Omega^2(P,\mathfrak{g})$ associated to a connection 1-form in a $G$-principal bundle represent the force field that you want to describe (for example $\Omega$ represents the electromagnetic field if $G=U(1)$, the strong force if $G=SU(3)$...) and so should be a physical quantity. But if you perform a gauge transformation, the curvature 2-form $\Omega$ change as $$ \Omega \rightarrow \Omega'=Ad_{g^{-1}} \circ \Omega$$

Question : Does 2-form curvature represent a physical quantity? If not, what is the object that represents the force fields?

$\endgroup$
1
  • 1
    $\begingroup$ The curvature does not change if $G=\mathrm{U}(1)$, since the adjoint of Abelian groups is trivial and the relation between the electromagnetic fields and the curvature tensor is an elementary part of the covariant formulation of electromagnetism, and is given explicitly e.g. on Wikipedia. What sources are you using that teach you about "curvature 2-forms" in gauge theory but not about their basic relations to observable physics? $\endgroup$
    – ACuriousMind
    Commented Jan 31 at 20:33

1 Answer 1

1
$\begingroup$
  1. For Abelian gauge groups, the adjoint representation is trivial so the curvature $F$ is itself gauge-invariant.

  2. For non-Abelian groups, gauge-invariant local operators are given by Ad-invariant polynomial functions of the curvature, e.g. the integrand $\mathrm{tr}_\mathfrak{g}(F\wedge{\star}F)$, where $\mathrm{tr}_\mathfrak{g}$ is the trace in the adjoint also used in the construction of the Killing form. Non-local gauge invariant operators are the Wilson loops, i.e. the holonomies of the gauge field.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.