"Regularity Singularities" and the Scattering of Gravity Waves in Approximate Locally Inertial Frames

TL;DR

This paper investigates whether regularity singularities exist at shock wave interactions in Einstein-Euler solutions, identifying geometric effects that may be physically real or removable by coordinate transformations.

Contribution

The authors identify new Coriolis-type effects in shock wave metrics that are essential and cannot be eliminated by smooth coordinate changes, clarifying the nature of regularity singularities.

Findings

Coriolis-type effects are fundamental in shock wave metrics.

These effects cannot be removed by smooth coordinate transformations.

The existence of regularity singularities relates to whether these effects are physically real.

Abstract

It is an open question whether solutions of the Einstein-Euler equations are smooth enough to admit locally inertial coordinates at points of shock wave interaction, or whether "regularity singularities" can exist at such points. The term {\it regularity singularity} was proposed by the authors as a point in spacetime where the gravitational metric tensor is Lipschitz continuous ($C^{0,1}$), but no smoother, in any coordinate system of the $C^{1,1}$ atlas. An existence theory for shock wave solutions in $C^{0,1}$ admitting arbitrary interactions has been proven for the Einstein-Euler equations in spherically symmetric spacetimes, but $C^{1,1}$ is the requisite smoothness required for space-time to be locally flat. Thus the open problem of regularity singularities is the problem as to whether locally inertial coordinate systems exist at shock waves within the larger $C^{1,1}$ atlas. To clarify this open problem, we identify new "Coriolis type" effects in the geometry of $C^{0,1}$ shock wave metrics and prove they are essential in the sense that they can never be made to vanish within the atlas of {\it smooth} coordinate transformations, the atlas usually assumed in classical differential geometry. Thus the problem of existence of regularity singularities is equivalent to the question as to whether or not these Coriolis type effects are essentially non-removable and `real', or merely coordinate effects that can be removed, (in analogy to classical Coriolis forces), by going to the less regular atlas of $C^{1,1}$ transformations. If essentially non-removable, it would argue strongly for a `real' new physical effect for General Relativity, providing a physical context to the open problem of regularity singularities.