10
$\begingroup$

On every Riemannian manifold $M$, we can consider the Hodge $*$-operator, which is characterised by the following formula: $$a \wedge *b = (a,b)\nu.$$ Here $a$ and $b$ are smooth forms on $M$, $(\ ,\ )$ is a metric on $\wedge T^*\!M$ and $\nu$ is the volume form with respect to the Riemannian metric.

My question: Is a formula of $*(a \wedge b)$ known?

I suspect that we can have a formula like "$*(a \wedge b)=(*a)\wedge(*b)$" or "$*(a \wedge b) = *a \wedge b \pm a \wedge *b$". Of course these formulae never hold. (Look at the degree.)

$\endgroup$
  • 2
    $\begingroup$ I put an answer to your question in the MO thread: mathoverflow.net/questions/162366/… $\endgroup$ – Ryan Budney Mar 25 '15 at 23:00
  • $\begingroup$ @RyanBudney Thank you. Your formula is simpler than I expected, even though the contraction is non-trivial. $\endgroup$ – H. Shindoh Mar 26 '15 at 10:56
2
$\begingroup$

A toy example. Given an orthonormal basis $e_1,\dots,e_n$ for the vector space $V$, then $\nu=e_1\wedge\dots\wedge e_n$ and

$$*(e_1\wedge e_2)=e_3\wedge\dots\wedge e_n\in \bigwedge^{n-2}V, $$

with $*e_1=e_2\wedge\dots\wedge e_n$ and $*e_2=-e_1\wedge e_3\wedge\dots\wedge e_n$ both in $\bigwedge^{n-1}V$. Introducing the insertion operator $i_\bullet: \bigwedge^{k}V\rightarrow \bigwedge^{k-1}V$ we arrive at

$$*(e_1\wedge e_2)=\frac{1}{2}\left( \underbrace{i_{e_2}(\underbrace{*e_1}_{\in \bigwedge^{n-1}V}}_{\in \bigwedge^{n-2}V})-i_{e_1}(*e_2)\right). $$

I do not if this is what you were searching for, but I hope it helps (a bit).

$\endgroup$
  • $\begingroup$ For related formulae, I refer to the "Three dimensional example" with $\star (\mathbf u \wedge \mathbf v )=\mathbf {u \times v}\,,\quad\star (\mathbf u \times \mathbf v ) = \mathbf u \wedge \mathbf v \,,$ in en.wikipedia.org/wiki/Hodge_dual $\endgroup$ – Avitus Dec 13 '13 at 20:33
  • 1
    $\begingroup$ Thanks for the formula. It could suggest that we can not have a simple formula (or may need a special notation). $\endgroup$ – H. Shindoh Dec 15 '13 at 20:26
  • $\begingroup$ You are welcome. No "simple" formula (like Jacobi identity, as you pointed out) is probably true. I found interesting the "dual" relationship between contraction and Hodge, though. The 3-dim case in the above comment is something non trivial. $\endgroup$ – Avitus Dec 15 '13 at 20:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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