The impact of initial conditions on quasi-normal modes

This study investigates the influence of initial conditions on the evolution and properties of linear quasi-normal modes (QNMs). Using a toy model with a delta effective potential in which the quasi-normal mode can be unambiguously identified, we highlight an aspect of QNMs that is long known yet often ignored: the amplitude of a QNM (after factoring out the corresponding exponential with a complex frequency) is not constant but instead varies with time. We stress that this is true even within the regime of validity of linear perturbation theory. The precise time variation depends on the initial conditions, as we show here for Gaussian and S-shaped initial conditions. Furthermore, it is possible to find initial conditions for which the QNM even fails to materialize; it is also possible to find those for which the QNM amplitude grows indefinitely. Focusing on cases where the QNM amplitude does stabilize at late times, we explore how this timescale for amplitude stabilization depends on the shape and location of the initial perturbation profile. Our findings underscore the need for care in fitting linear QNMs to ringdown data, and for a better understanding of the initial conditions. They also suggest recent computations of quadratic QNMs, sourced purely by stabilized linear QNMs, may not fully capture what determines the amplitude of the quadratic QNMs, even at late times. Our results motivate a detailed investigation of the initial perturbations generated in the aftermath of a binary merger.