Let $M$ be a compact manifold (without boundary) and let $f:M\to \mathbb{R}$ be a fixed Morse-function.

Question: Is the space $GVect(M,f)$ of all gradient-like vector-fields for $f$ contractible or even convex?

To eliminate any misunderstanding, this is the definition of gradient-like vector fields that I'm using:

Definition: A vector field $X\in Vect(M)$ is called gradient-like for $f$ if
1. $\forall q\in M\setminus Crit(f):\; df(q)X(q)<0$.
2. For each critical point $p\in Crit(f)$ of $f$, $\exists$ Morse coordinate chart $\phi:U\to\mathbb{R^n}$ (i.e. $\phi(p)=0$ and $f\circ\phi^{-1}(x)=f(p)-\sum_{i=1}^k x_i^2+\sum_{i=k+1}^n x_i^2$ for $x\in\phi(U)$, $k=index(p)$) such that $$\phi_*X(x)=d\phi(\phi^{-1}(x))X(\phi^{-1}(x))=(2x_1,\ldots,2x_k,-2x_{k+1},\ldots,-2x_n)^T$$ for $x\in\phi(U)$ (i.e. in this Morse-chart $X$ coincides with the negative gradient of $f$ w.r.t. the euclidean metric on $\mathbb{R^n}$).

Thank you in advance for any insights.


$GVect(M,f)$ is a convex set in the space of all vector fields. Thus it is contractible.

Edit, since you asked for details: I assume that the Morse charts $\Phi$ are the same for all $X$.
Let $X,Y$ be gradient-like for $f$, then:
1. $df(q)(t.X(q)+(1-t).Y(q)) = t.df(q)(X(q)) + (1-t).df(q)(Y(q)) < 0$ for $0\le t\le 1$.
2. In the Morse charts the two vector fields look the same.

Second edit: Okay, you want the Morse chart to depend on the vector field. Maybe, your definition of a gradient-like vector field is equivalent to:

a) At a critical point $p$ we have $X(p)=0$.
b) Near each critical point $p$, the function $-X(f)= -df(X)$ is Morse with critical point $p$ a minimum. Then you can use $-df(X)$ as the square distance from $p$ for a Riemannian metric, and the Morse chart construction with respect to that Riemannian metric leads to a Morse chart as you require.
c) Off the critical points, we have $df(X)<0$ (i.e. X is transversal to the level sets and points towards lower level sets).

  • 1
    $\begingroup$ Could you please elaborate? Why exactly is it convex? $\endgroup$
    – Dave
    Mar 8 '13 at 16:15
  • $\begingroup$ Why can you assume that the Morse charts are the same for the two gradient-like vector fields? This is not clear to me at all. $\endgroup$
    – Dave
    Mar 8 '13 at 17:11
  • $\begingroup$ You did not say that they depend on the field. Actually, into the construction of a Morse chart goes a Riemann metric and the neg. gradient of $f$ with respect to $g$. The space of Riemann metrics is also convex. $\endgroup$ Mar 8 '13 at 17:28
  • $\begingroup$ How much freedom do you have in choosing a Morse chart for a fixed Morse function? Maybe, this is the core of your problem! $\endgroup$ Mar 8 '13 at 17:34
  • 2
    $\begingroup$ There are lots of Morse charts around a Morse point. Any diffeomorphism preserving the level sets will give you another Morse chart, so you can ``slide around'' the level sets. For example, in the case of a positive definite Morse point, you can rotate the spherical level sets, each by a different angle. $\endgroup$
    – Ben McKay
    Mar 11 '13 at 17:06

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.