Spectro-timing study of GX 339-4 in a hard intermediate state

TL;DR

This study analyzes NuSTAR data of GX 339-4 during a hard intermediate state, revealing spectral and timing evolution, including a correlation between QPO frequency and spectral softening, challenging existing models of accretion physics.

Contribution

It provides a detailed spectral and timing analysis of GX 339-4 in a specific state, highlighting discrepancies with current theoretical models of QPOs and accretion disk structure.

Findings

QPO frequency increases from 0.68 to 1.05 Hz as spectrum softens.

Inner disk radius remains around 9 gravitational radii throughout the observation.

Current models cannot fully explain the observed QPOs and disk radii without assuming an unlikely black hole mass.

Abstract

We present an analysis of NuSTAR observations of a hard intermediate state of the transient black hole GX 339-4 taken in January 2015. As the source softened significantly over the course of the 1.3 d-long observation we split the data into 21 sub-sets and find that the spectrum of all of them can be well described by a power-law continuum with an additional relativistically blurred reflection component. The photon index increases from ~1.69 to ~1.77 over the course of the observation. The accretion disk is truncated at around 9 gravitational radii in all spectra. We also perform timing analysis on the same 21 individual data sets, and find a strong type-C quasi-periodic oscillation (QPO), which increase in frequency from ~0.68 to ~1.05 Hz with time. The frequency change is well correlated with the softening of the spectrum. We discuss possible scenarios for the production of the QPO and calculate predicted inner radii in the relativistic precession model as well as the global disk mode oscillations model. We find discrepancies with respect to the observed values in both models unless we allow for a black hole mass of ~100 M_sun , which is highly unlikely. We discuss possible systematic uncertainties, in particular with the measurement of the inner accretion disk radius in the relativistic reflection model. We conclude that the combination of observed QPO frequencies and inner accretion disk radii, as obtained from spectral fitting, is difficult to reconcile with current models.