Multi-timescale X-ray reverberation mapping of accreting black holes

TL;DR

This paper introduces a novel reverberation mapping formalism for accreting black holes that models energy-dependent time lags across multiple timescales, accounting for spectral variability and non-linear effects, enabling more accurate measurements.

Contribution

It presents an analytic reverberation mapping model that incorporates spectral variability and non-linear effects, allowing simultaneous fitting over a wide frequency range with improved accuracy.

Findings

Bias in lag measurements when ignoring non-linear effects

Successful fits to Cygnus X-1 data demonstrating model effectiveness

Enhanced understanding of accretion disk reflection features

Abstract

Accreting black holes show characteristic reflection features in their X-ray spectrum, including an iron K$\alpha$ line, resulting from hard X-ray continuum photons illuminating the accretion disk. The reverberation lag resulting from the path length difference between direct and reflected emission provides a powerful tool to probe the innermost regions around both stellar-mass and supermassive black holes. Here, we present for the first time a reverberation mapping formalism that enables modeling of energy dependent time lags and variability amplitude for a wide range of variability timescales, taking the complete information of the cross-spectrum into account. We use a pivoting power-law model to account for the spectral variability of the continuum that dominates over the reverberation lags for longer time scale variability. We use an analytic approximation to self-consistently account for the non-linear effects caused by this continuum spectral variability, which have been ignored by all previous reverberation studies. We find that ignoring these non-linear effects can bias measurements of the reverberation lags, particularly at low frequencies. Since our model is analytic, we are able to fit simultaneously for a wide range of Fourier frequencies without prohibitive computational expense. We also introduce a formalism of fitting to real and imaginary parts of our cross-spectrum statistic, which naturally avoids some mistakes/inaccuracies previously common in the literature. We perform proof-of-principle fits to Rossi X-ray Timing Explorer data of Cygnus X-1.