ProbXiv
Explorersign in
Problem archiveProblem record

Statement

Problem 12.0.5. Construct a system of scalar invariants of a projective structure that vanish if and only if there exists (locally, in a neighborhood of almost every point) a metric compatible with a given projective structure.

In dimension two, the problem was solved in [Br-Du-Eas2009], and the system of invariants is quite complicated - the simplest invariant has degree 5 in derivatives. It is possible that in higher dimensions the system of invariants could be easier in some ways, since in this case the PDE-system corresponding to the existence of a metric for a projective structure has a higher degree of overdeterminacy. In particular fewer differentiations are needed to construct the first obstruction to the existence of a metric class. See the recent paper [Dun-Eas2014]. Let us now discuss the freedom in reconstructing the metric by its geodesics; an equivalent question is how many different metrics can have the same geodesics considered as unparameterized curves. Of course, the metrics g and const · g have the same geodesics; in [Matveev2012a] it was shown that for a generic metric any projectively equivalent metric is proportional to it with a constant coefficient. There exist local (the first examples are due already to Lagrange, Beltrami and Dini) and global examples of nonproportional projectively equivalent metrics. Locally, in the Riemannian case, a complete description of projectively equivalent metrics is due to Levi-Civita [Levi-Civita1896] and in arbitrary signature is due to [Bol-Mat2013]. Globally, in the Riemannian case, the situation is also pretty clear, but if the metrics have arbitrary signature, virtually nothing is known. A general problem is to understand topology of closed manifolds admitting nonproptional projectively equivalent metrics of indefinite signature and as the simplest version of this problem we suggest

Record

Added
Links

Comments

Nothing has been published against this problem yet, and nothing has been said about it.

Sign in with an institutional address to take part in the discussion. Reading every thread stays open to everyone.

Sign in

Solve with an agent

Open the statement in a chat, with the problem and the ground rules already written into the prompt.

This opens a third-party site. Nothing is posted back to ProbXiv and nothing you write there is recorded here — what a model gives you is an attempt, which a person still has to check.