I'm stuck with the following claim made in Section 13.1 of Y-G. Oh's book "Symplectic topology and Floer homology". Assume that $N$ is a differential manifold and $S_0 ,S_1\subseteq N$ two submanifolds of $N$, and we consider the corresponding conormal bundles $\nu^{\ast} S_0$ and $\nu^{\ast} S_1$. Let $H\colon T^{\ast} N\times I\to\mathbb{R}$ be a smooth function and $X_H$ the corresponding Hamiltonian vector field, and $\phi_H^t$ the corresponding Hamiltonian flow. We define a function $g\colon\nu^{\ast} S_1\to\mathbb{R}$ by $$ g(x)=\mathcal{A}_H (z_H^x), $$ where $\mathcal{A}_H$ is the corresponding action functional, and $z_H^x$ the path given by $z_H^x (t)=\phi_H^t\circ(\phi_H^1)^{-1}(x)$. Then he claim that:

  1. The intersection $\phi_H^1 (\nu^{\ast} S_0)\cap\nu^{\ast} S_1$ is compact in $T^{\ast} N$.

It's obviously not the case: if we pick $H=0$, then the condition is equivalent to saying that $S_0$ and $S_1$ are disjoint. But this is a procedure to his proof of the fact that the action spectrum is compact and has measure zero. So I'm wondering if

1'. The image $g(\phi_H^1 (\nu^{\ast} S_0)\cap\nu^{\ast} S_1)$ is compact in $\mathbb{R}$.

is true.

Another claim is about the linearization of the ODO $\frac{\partial }{\partial t} -X_H$, namely $\nabla_{\dot{\gamma }} -DX_H (\gamma)$. We consider the path space with boundary conditions $\Omega_{S_0 S_1} (T^{\ast} N,\mathrm{d}\lambda)$, given by all paths starting from $\nu^{\ast} S_0$ and ends at $\nu^{\ast} S_1$. Then the tangent space of this path space at a path $\gamma$ is given by $\Gamma_{S_0S_1} (\gamma^{\ast} (TT^{\ast} N))$, the space of vector fields on $\gamma$ with endpoints in the given tangent spaces. The linearization is then a linear operator $\nabla_{\dot{\gamma}} -\mathrm{D} X_H\colon\Gamma_{S_0S_1} (\gamma^{\ast} (TT^{\ast}N))\to\Gamma (\gamma^{\ast} (TT^{\ast} N))$. Then the book claims that

  1. The intersection $\phi_H^1 (\nu^{\ast} S_0)\cap\nu^{\ast} S_1$ is transversal if and only if the linearization $\nabla_{\dot{\gamma }} -\mathrm{D}X_H$ is surjective.

I have no idea why this is true. I don't know how to relate the surjectivity of the linearization to the transversality of intersections. Any comments or answers are helpful.


Your Answer

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

Browse other questions tagged or ask your own question.