[Paper Review] Selecting Very Soft X-Ray Sources in External Galaxies: Luminous Supersoft X-ray Sources and Quasisoft Sources

R. Di Stefano, A. K. H. Kong|arXiv (Cornell University)|Nov 17, 2003

Galaxies: Formation, Evolution, Phenomena2 references3 citations

TL;DR

This paper presents a systematic algorithm to identify very soft X-ray sources (VSSs), including luminous supersoft sources (SSSs) and quasisoft sources (QSSs), in external galaxies using Chandra and XMM-Newton data. The method applies spectral fitting criteria—such as blackbody temperature <175 eV or power-law photon index >3.5—to select SSSs and QSSs, achieving 90% SSS recovery and minimal contamination from hard sources, with QSSs likely representing accreting neutron stars or intermediate-mass black holes.

ABSTRACT

We introduce a procedure to identify very soft X-ray sources (VSSs) in external galaxies. Our immediate goal was to formulate a systematic procedure to identify luminous supersoft X-ray sources (SSSs), so as to allow comparisons among galaxies and to study environmental effects. The focus of this paper is on the design of the selection algorithm and on its application to simulated data. In the companion paper we test it by applying it to sources discovered through Chandra observations of 4 galaxies. We find that, in its application to both simulated and real data, our procedure also selects somewhat harder sources, which we call quasisoft. Whereas values of kT for SSSs are typically tens of eV, some quasisoft sources (QSSs) may have kT as high as ~250-300 eV. The dominant spectral component of other QSSs may be as soft as SSS spectra, but the spectra may also include a low-luminosity harder component. We sketch physical models for both supersoft and quasisoft sources. Some SSSs are likely to be accreting white dwarfs; some of these may be progenitors of Type Ia supernovae. Most QSSs may be too hot to be white dwarfs. They, together with a subset of SSSs, may be neutron stars or, perhaps most likely, accreting intermediate-mass black holes.

Motivation & Objective

• To develop a well-defined, systematic procedure for identifying luminous supersoft X-ray sources (SSSs) in external galaxies, especially where spectral fitting is limited by low count rates.
• To extend the selection to slightly harder sources (quasisoft sources, QSSs) that may be physically distinct from canonical SSSs but are often missed by traditional criteria.
• To enable comparative studies of SSS populations across galaxies and to probe environmental and galactic effects on these sources.
• To validate the method using simulated data and prepare for application to real Chandra and XMM-Newton observations.
• To distinguish between true SSSs and QSSs based on algorithmic conditions, particularly the HR and 3σ criteria, to improve source classification.

Proposed method

• The algorithm uses a combination of hardness ratio (HR) and significance-based (3σ) criteria to identify sources likely to have very soft spectra, even with low count rates.
• It applies three main selection conditions: the HR condition, the 3σ condition, and a weaker condition for quasisoft sources (QSSs), with SSSs defined as satisfying either the HR or 3σ condition.
• The method is calibrated using simulated X-ray data with known spectral parameters, including blackbody models with kT < 175 eV and power-law models with α > 3.5.
• Sources are classified as SSSs if their inferred spectral parameters meet the strict criteria; those satisfying only the weaker conditions are labeled QSSs.
• The algorithm is designed to be robust across both high- and low-count-rate sources, ensuring consistent application to Chandra and XMM-Newton data.
• It incorporates a spectral energy cutoff criterion: less than 10% of the total energy in photons with E > 1.5 keV, to identify very soft emission.

Experimental results

Research questions

• RQ1 How can we systematically identify luminous supersoft X-ray sources (SSSs) in external galaxies when spectral fitting is often impossible due to low count rates?
• RQ2 What is the nature and physical origin of sources that are slightly harder than canonical SSSs, and how can they be distinguished from true SSSs?
• RQ3 To what extent does the selection algorithm recover known SSSs in simulated data, and how much contamination from hard sources does it introduce?
• RQ4 Are quasisoft sources (QSSs) physically distinct from SSSs, and what astrophysical models might explain their existence?
• RQ5 Can the algorithm be reliably applied to real Chandra and XMM-Newton data to identify SSS and QSS populations in multiple galaxies?

Key findings

• The algorithm successfully identifies 90% of simulated blackbody SSSs with kT < 175 eV, demonstrating high sensitivity to the target source population.
• Only 4% of simulated hard sources (kT > 500 eV) are misclassified as SSSs, indicating low false-positive contamination.
• The method recovers 70% of quasisoft sources (QSSs) with kT in the range 175–500 eV, confirming its ability to detect harder but still very soft sources.
• The selection process inevitably includes QSSs with kT up to ~250–300 eV or with a soft primary component and a low-luminosity hard component, suggesting a broader class of very soft sources.
• QSSs are likely not white dwarfs, but may instead be accreting neutron stars or intermediate-mass black holes, based on their spectral hardness and luminosity.
• The algorithm performs robustly across both high- and low-count regimes, enabling consistent application to real Chandra and XMM-Newton data.