1
$\begingroup$

Consider $M$ a smooth compact connected manifold with $w$ a volum form. Take for example $M=\mathbb{S}^n$ with the uniform measure. For any smooth map $f:M\rightarrow M$ and its pullback measure $\nu = f^* w$ it is well known that we have $$\int_M \nu = \text{deg}(f)\int_M w $$with $\deg(f)\in \mathbb{Z}$. Is this condition sufficient? For any $\nu$ volum form on $M$ such that $\int_M \nu / \int_M w \in \mathbb{Z} $ does there exist $f:M\rightarrow M$ such that $\nu = f^*w$ ? Can we construct such an $f$ explicitly ?

$\endgroup$
2
  • 4
    $\begingroup$ $S^n$ does not have a Haar measure unless $n=0,1,3$. Maybe $7$ if you stretch the meaning of Haar measure a bit. Ignoring this, most manifolds do not have self-maps of arbitrary degree; and if you want $\nu$ to also be nonvanishing then $f$ should be a covering map (even less common). It is a theorem of Moser that if $\int \omega = \int \nu$ for two nonvanishing top forms, then there's an oriented diffeomorphism $f$ with $f^*\nu = \omega$. The proof is what's usually called the Moser trick. $\endgroup$
    – mme
    Sep 3 '20 at 13:09
  • $\begingroup$ a note on terminology: usually measures are only pushed forward, and differential forms are pulled back. Defining pullback measures is tricky mathoverflow.net/q/122704/3948 . $\endgroup$ Sep 3 '20 at 16:08
3
$\begingroup$

No. Consider the $2$-sphere, and the standard volume $\omega$ form with volume $4\pi.$ Consider $\nu=2\omega.$ If $f\colon S^2\to S^2$ with $f^*\omega=\nu$ would exist, it would be a local diffeomorphism and by compactness a covering map. As $S^2$ is simply connected and $f$ would have degree 2, this is not possible.

$\endgroup$

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.