Abstract:

Be it numerical, formal, or rigorous, the typical setup for the evolution of spacetimes describing, for instance, the merger of two black holes is to make an ad hoc assumption on the initial data to have compactly supported gravitational radiation content.

Informally, this ensures that the spacetime has a smooth null infinity (it "peels"), but essentially pretends the spacetime has been stationary at some time "before" the initial data.
My talk will be about what happens when one does not ignore the past dynamics of the system: I will present mathematical results that convert scattering data assumptions in the infinite past into precise predictions on the asymptotics towards future null infinity, painting a detailed picture of the asymptotics of physically realistic spacetimes at early times which, in particular, does not leave room for a smooth null infinity.

Furthermore, I will explain how this failure of null infinity to be smooth affects the asymptotics of gravitational radiation at late times ("tails"), predicting decay up to three powers slower than the usual "Price's law" decay.