Ehrhart polynomials and symplectic embeddings of ellipsoids

TL;DR

This paper explores symplectic embeddings of four-dimensional ellipsoids, revealing new infinite staircases and connecting these phenomena to Ehrhart quasipolynomial properties, with implications for embedding obstructions.

Contribution

The authors provide new proofs of known staircases and discover a new staircase for embeddings into E(1,3/2), linking symplectic embedding obstructions to Ehrhart quasipolynomial phenomena.

Findings

New proofs of McDuff-Schlenk and Frenkel-Müller staircases

Discovery of a new staircase for E(1,3/2) embeddings

Volume is the only obstruction for large a in E(1,a) embeddings

Abstract

McDuff and Schlenk determined when a four-dimensional ellipsoid can be symplectically embedded into a ball, and found that part of the answer is given by an infinite "Fibonacci staircase." Similarly, Frenkel and M\"uller determined when a four-dimensional ellipsoid can be symplectically embedded into the ellipsoid E(1,2) and found that part of the answer is given by a "Pell staircase." ECH capacities give an obstruction to symplectically embedding one four-dimensional ellipsoid into another, and McDuff showed that this obstruction is sharp. We use this result to give new proofs of the staircases of McDuff-Schlenk and Frenkel-M\"uller, and we prove that another infinite staircase arises for embeddings into the ellipsoid E(1,3/2). Our proofs relate these staircases to a combinatorial phenomenon of independent interest called "period collapse" of the Ehrhart quasipolynomial. In the appendix, we use McDuff's theorem to show that for a >= 6, the only obstruction to embedding an ellipsoid E(1,a) into a scaling of E(1,3/2) is the volume, and we also give new proofs of similar results for embeddings into scalings of E(1,1) and E(1,2).