[Paper Review] Simultaneous independent measurements of a truncated inner accretion disc in the low/hard state of GX 399$-$4

Daniel Plant, K. O’Brien+1 more|arXiv (Cornell University)|Nov 26, 2014

Astrophysical Phenomena and Observations2 references3 citations

TL;DR

This study presents the first simultaneous measurement of the inner accretion disc radius in GX 339-4's low/hard state using XMM-Newton EPIC-pn imaging mode, achieving unprecedented spectral calibration down to 0.4 keV. Both the thermal disc component and the broadened Fe Kα line independently yield consistent inner radii of 20–30 r₉, providing strong evidence for a truncated disc in the hard state.

ABSTRACT

We present results from three recent XMM-Newton observations of GX 339$-$4 in the low/hard state, taken during the decay of a bright (peak $\sim 0.05$ L$_{ m Edd}$) failed outburst. Uniquely, these are the first XMM-Newton EPIC-pn observations of this source using an imaging mode, which significantly enhances the quality of the data at hand. In particular, thanks to the larger available bandpass, this allows an unprecedented constraint of the thermal accretion disc component, and the level of photoelectric absorption. We simultaneously measured the inner radius of the accretion disc via the broadened Fe K$α$ line and the disc component. The two methods agree, and the measured radii show good consistency over the three epochs. We find that the inner radius is at 20-30 $r_{ m g}$, adding to the growing direct evidence for truncation of the inner accretion disc in the low/hard state.

Motivation & Objective

• To resolve the long-standing debate on whether the accretion disc is truncated or extends to the ISCO in the low/hard state of black hole X-ray binaries.
• To improve spectral calibration and constrain the thermal disc component by using XMM-Newton EPIC-pn in imaging mode, which extends sensitivity down to 0.4 keV.
• To perform simultaneous, independent measurements of the inner disc radius using both the thermal disc component and the relativistically broadened Fe Kα line.
• To test the consistency of disc and reflection-based inner radius estimates in the hard state, providing a stringent test of the truncated disc model.

Proposed method

• Utilized three XMM-Newton EPIC-pn observations of GX 339-4 in small window science mode, enabling improved spectral calibration down to 0.4 keV compared to standard fast modes.
• Applied the relxilllp reflection model to self-consistently fit the power-law and reflection components, including source height and emissivity profile as free parameters.
• Simultaneously fitted the thermal disc component (using a multicolor disc blackbody model) and the Fe Kα line profile to derive the inner disc radius from both methods.
• Used MCMC sampling to derive 90% confidence intervals for all parameters, including inclination, disc inner radius, and source height.
• Relaxed the standard R⁻³ emissivity assumption by using relxilllp, which models the illuminating source height and corresponding emissivity, improving physical consistency.
• Corrected for pile-up using the XMM-Newton standard reduction pipeline and applied the most recent calibration files (CCFs) to ensure spectral accuracy.

Experimental results

Research questions

• RQ1 Is the inner accretion disc truncated in the low/hard state of GX 339-4, as predicted by ADAF models?
• RQ2 Do independent measurements of the inner disc radius from the thermal disc component and the relativistically broadened Fe Kα line yield consistent results in the hard state?
• RQ3 How does the use of XMM-Newton EPIC-pn imaging mode, with extended low-energy bandpass, improve constraints on the disc component and inner radius?
• RQ4 What is the impact of inclination on the derived inner radius from both disc and reflection components, and how does it affect consistency between methods?

Key findings

• The thermal disc component, fitted via a multicolor disc blackbody model, yields an inner radius of 25 r₉ (±2) in the first observation, 29 r₉ (±3) in the second, and 26 r₉ (+3/−1) in the third.
• The broadened Fe Kα line component yields an inner radius of 20 r₉ (+4/−5) in the first observation, 29 r₉ (±2) in the second, and 16 r₉ (±2) in the third, showing good agreement with the disc-based estimates.
• The joint fit with the relxilllp model yields a source height of 14–70 r₉ (90% confidence), with poor constraints due to degeneracy in emissivity profiles at these radii.
• The fitted inclination is 33° (±3°), which is lower than expected from mass function constraints, and this lower inclination contributes to the observed smaller inner radius estimates.
• Using the relxilllp model instead of R⁻³ emissivity improves the fit by Δχ² = −21 and maintains consistency in the inner radius estimates, confirming the robustness of the results.
• This study provides the first simultaneous, consistent measurement of disc truncation in the hard state using both disc and reflection-based methods, strengthening the evidence for a truncated disc in low/hard states.