Reconstruction of the Alcubierre shaping function from simulated data

Abstract

Given noisy simulated spacetime-curvature measurements around an Alcubierre bubble, we demonstrate that the shaping function f(r) can be reconstructed with <5% RMS error using Bayesian inversion with sparse Gaussian-process priors. This establishes the feasibility of remote diagnostics of propulsive spacetime geometries, should any ever be produced.