Scalar-field amplitudes in black-hole evaporation

Abstract

We consider the quantum-mechanical decay of a Schwarzschild-like black hole into almost-flat space and weak radiation at a very late time. That is, we are concerned with evaluating quantum amplitudes (not just probabilities) for transitions from initial to final states. In this quantum description, no information is lost because of the black hole. The Lagrangian is taken, in the first instance, to consist of the simplest locally supersymmetric generalization of Einstein gravity and a massless scalar field. The quantum amplitude to go from given initial to final bosonic data in a slightly complexified time-interval T=τexp(−iθ) at infinity may be approximated by the form const×exp(−I), where I is the (complex) Euclidean action of the classical solution filling in between the boundary data. Additionally, in a pure supergravity theory, the amplitude const×exp(−I) is exact. Suppose that Dirichlet boundary data for gravity and the scalar field are posed on an initial spacelike hypersurface extending to spatial infinity, just prior to collapse, and on a corresponding final spacelike surface, sufficiently far to the future of the initial surface to catch all the Hawking radiation. Only in an averaged sense will this radiation have an approximately spherically-symmetric distribution. If the time-interval T had been taken to be exactly real, then the resulting ‘hyperbolic Dirichlet boundary-value problem’ would, as is well known, not be well posed. Provided instead (‘Euclidean strategy’) that one takes T complex, as above (0<θ⩽π/2), one expects that the field equations become strongly elliptic, and that there exists a unique solution to the classical boundary-value problem. Within this context, by expanding the bosonic part of the action to quadratic order in perturbations about the classical solution, one obtains the quantum amplitude for weak-field final configurations, up to normalization. Such amplitudes are here calculated for weak final scalar fields.