Systematic Analysis of Low/Hard State RXTE Spectra of GX 339-4 to Constrain the Geometry of the System

TL;DR

This study analyzes RXTE spectra of GX 339-4 in the low/hard state to investigate whether the accretion disk extends to the innermost stable orbit, reconciling broad Fe line presence with spectral hardness.

Contribution

It provides a systematic analysis of RXTE data to test if reflection components can explain the spectral properties in the hard state of GX 339-4.

Findings

Some observations show increased photon index with reflection inclusion.

Broad Fe lines are present even when photon index is less than 2.

The results challenge the simple truncated disk model for the hard state.

Abstract

One of the popular models for the low/hard state of Black Hole Binaries is that the standard accretion disk is truncated and the hot inner region produces via Comptonization, the hard X-ray flux. This is supported by the value of the high energy photon index, which is often found to be small $\sim$ 1.7 ($<$ 2) implying that the hot medium is seed photons starved. On the other hand, the suggestive presence of a broad relativistic Fe line during the hard state would suggest that the accretion disk is not truncated but extends all the way to the inner most stable circle orbit. In such a case, it is a puzzle why the hot medium would remain photon starved. The broad Fe line should be accompanied by a broad smeared reflection hump at $\sim$ 30 keV and it may be that this additional component makes the spectrum hard and the intrinsic photon index is larger, i.e. $>$ 2. This would mean that the medium is not photon deficient, reconciling the presence of a broad Fe line in the observed hard state. To test this hypothesis, we have analyzed the RXTE observations of GX 339-4 from the four outbursts during 2002-2011 and identify the observations when the system was in the hard state and showed a broad Fe line. We have then attempted to fit these observations with models, which include smeared reflection to understand whether the intrinsic photon index can indeed be large. We find that, while for some observations the inclusion of reflection does increase the photon index, there are hard state observations with broad Fe line that have photon indices less than 2.