3
$\begingroup$

Let $S^4$ be a round sphere of radius 1 (with the standard Riemannian round metric), and let $D_\text{F} \equiv \gamma^\mu \nabla_\mu$ be the Dirac operator on $S^4$, acting on the usual spinors for the Riemannian metric.

I wonder what is its Green's function $G_\text{F}$ on $S^4$, given by the equation $ D_\text{F} G_\text{F}(x)_{\alpha \beta} = \delta_{S^4}(x) \delta_{\alpha \beta}?$

I happen to know the Green's function for the conformally coupled scalar, namely Green's function $G_\text{B}$ for $$\Delta_\text{scalar} \equiv - \nabla^\mu \nabla_\mu + 2$$ acting on scalar functions. But I think $\gamma^\mu \nabla_\mu \gamma^\nu \nabla_\nu$ slightly differs from $\Delta_\text{scalar}$ by a constant proportional to the $S^4$ scalar curvature, so it's not obvious to me how to get $G_\text{F}$ from $G_\text{B}$ (which we do all the time on flat space).

So more generally, I'd like to know the Green's function for the Dirac operator on $S^n$.

References and suggestions will be great!

$\endgroup$
2
  • $\begingroup$ There's no Lorentz metric on $S^4$ so I'm not sure you can really have a "Dirac equation", unless by this you mean for some spinor of the Riemannian metric of the 4-sphere $\endgroup$
    – Slereah
    Commented May 12, 2017 at 9:10
  • $\begingroup$ @Slereah Yes indeed I mean spinor for the Riemannian round metric. I hope this is not an obstruction to defining Green's function $\endgroup$
    – Lelouch
    Commented May 12, 2017 at 9:16

1 Answer 1

4
$\begingroup$

The spinor bundle on the standard sphere $S^n$ can be trivialized by Killing spinors as explained e.g. in an article by Baer. This means we have spinors $\psi_1,...,\psi_N$ such that $$ \nabla_X\psi_j=\mu X\cdot\psi_j $$ for all tangent vectors $X$ on $S^n$, where $\mu\in\{\pm\frac{1}{2}\}$, and every spinor $\Phi$ can be globally written as $\Phi=\sum_{j}f_j\psi_j$ with complex scalar functions $f_j$. Let $D$ be the Dirac operator. Analogously to Lemma 3 in the article mentioned above one gets $$ (D+\mu)^2\Phi=\sum_{j=1}^{N}\Big(\Delta_g f_j+\frac{(n-1)^2}{4}f_j\Big)\psi_j $$ and thus $$ D(D+2\mu)\Phi=\sum_{j=1}^{N}\Big(\Delta_g f_j+\frac{n(n-2)}{4}f_j\Big)\psi_j. $$ On the right hand side we have the conformal Laplace operator acting on the functions $f_j$. Therefore, if $f_j$ are multiples of the Green function for the conformal Laplace operator at a point $y\in S^n$, then $(D+2\mu)\Phi$ is a Green function for the Dirac operator at $y$.

$\endgroup$

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.