Abstract:A taming symplectic structure provides an upper bound on the area of an approximately pseudoholomorphic curve in terms of its homology class. We prove that, conversely, an almost complex manifold with such an area bound admits a taming symplectic structure. This confirms a speculation by Gromov. We also characterize the cone of taming symplectic structures numerically, prove that complex 2-cycles can be approximated by coarsely holomorphic curves, and provide a lower energy bound for such curves.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the relationship between the existence of rough holomorphic curves and symplectic structures in almost complex manifolds. Specifically, the main objective of the paper is to prove that if an almost complex manifold has a certain area upper bound condition, then this manifold must have a tame symplectic structure. This conclusion verifies one of Gromov's conjectures and provides new tools and methods for further research on geometric results in symplectic topology.
### Main contributions of the paper
1. **Main theorem**:
- **Theorem 1**: Let \((X, J)\) be a closed almost complex manifold, and \(h\in H^{2}(X,\mathbb{R})\) be a cohomology class. Then \(h\) contains a symplectic form that tames \(J\) if and only if there exist \(\epsilon > 0\) and \(C>0\) such that for every rough \(\epsilon\)-holomorphic curve \(S\subset X\), we have
\[
\text{area}(S)\leq C\langle h, [S]\rangle.
\]
2. **Numerical characterization of the symplectic cone**:
- **Corollary 1.6**: Let \((X, J)\) be a closed almost complex manifold, and \(J\) is tamed by some symplectic form. Then a cohomology class \(h\in H^{2}(X,\mathbb{R})\) contains a symplectic form that tames \(J\) if and only if there exists \(\epsilon > 0\) such that \(h\) is strictly positive on every rough \(\epsilon\)-holomorphic curve.
3. **Approximation of complex periods**:
- **Theorem 2**: Let \((X, J)\) be a closed almost complex manifold, and \(T\neq 0\) be a complex period. Then there exist sequences \(\epsilon_{i}>0\), \(c_{i}>0\) and \(S_{i}\subset X\) such that \(S_{i}\) is a rough \(\epsilon_{i}\)-holomorphic curve, \(\epsilon_{i}\) converges to zero, and \(c_{i}S_{i}\) converges to \(T\) in the weak topology.
4. **Lower bound of energy**:
- **Theorem 3**: Let \((X, J, g)\) be a compact almost complex Riemannian manifold. There exist \(\hbar > 0\) and \(\epsilon > 0\) such that for every rough \(\epsilon\)-holomorphic curve \(S\subset X\), we have
\[
\text{area}(S)>\hbar.
\]
### Methods for solving problems
- **Definition of rough holomorphic curves**: In the paper, rough \(\epsilon\)-holomorphic curves are defined, which are surfaces that are close to pseudo - holomorphic curves in an average sense.
- **Existence of symplectic structures**: By proving the existence of a tame symplectic structure, Gromov's conjecture is verified.
- **Approximation of complex periods**: Through approximation techniques, it is proved that complex periods can be approximated by rough holomorphic curves.
- **Lower bound of energy**: Using the regularity theory of semi - calibrated manifolds, the lower bound of the energy of rough holomorphic curves is proved.
### Conclusion
Through a series of strict mathematical proofs, this paper not only verifies Gromov's conjecture, but also provides a new research method for geometric results in symplectic topology. These results are of great significance for understanding the geometric properties of almost complex manifolds.