Burst-induced coronal cooling in GS 1826–24

Abstract

Type I X-ray bursts in GS 1826-24, and in several other systems, may induce cooling of the hot inner accretion flow that surrounds the bursting neutron star. Given that GS 1826-24 remained persistently in the hard state over the period 2003-2008 and presented regular bursting properties, we stacked the spectra of the X-ray bursts detected by INTEGRAL (JEM-X and ISGRI) and XMM-Newton (RGS) during that period to study the effect of the burst photons on the properties of the Comptonizing medium. The extended energy range provided by these instruments allows the simultaneous observation of the burst and persistent emission spectra. We detect an overall change in the shape of the persistent emission spectrum in response to the burst photon shower. For the first time, we observe simultaneously a drop in the hard X-ray emission, together with a soft X-ray excess with respect to the burst blackbody emission. The hard X-ray drop can be explained by burst-induced coronal cooling, while the bulk of the soft X-ray excess can be described by fitting the burst emission with an atmosphere model, instead of a simple blackbody model. Traditionally, the persistent emission was assumed to be invariant during X-ray bursts, and more recently to change only in normalization but not in spectral shape; the observed change in the persistent emission level during X-ray bursts may thus trigger the revision of existing neutron star massradius constraints, as the derived values rely on the assumption that the persistent emission does not change during X-ray bursts. The traditional burst fitting technique leads to up to a 10% overestimation of the bolometric burst flux in GS 1826-24, which significantly hampers the comparisons of the KEPLER and MESA model against this 'textbook burster'.

Cooling of accretion disc coronae by Type I X-ray bursts

“…However, the persistent spectrum produced by the accretion disc and corona will also radiate during the burst and must be accounted for in order to provide an accurate description of the burst. Several analyses of X-ray bursts with hard X-ray instruments have detected a hard X-ray shortage during the burst (e.g., Maccarone & Coppi 2003;Ji et al 2014a,b;Sánchez-Fernández et al 2020), implying that the persistent spectrum is affected by the burst, perhaps by Compton cooling due to the burst photons. This paper presents the results of a series of calculations that considered how burst-driven Compton cooling will effect the temperature and emitted spectrum of a thermal accretion disc corona.…”

“…1), and therefore simply multiplying the persistent spectrum by a factor during the burst does not accurately capture the physical processes occurring within the corona. Both changes in the normalization and spectral shape of the persistent spectrum should be considered when fitting X-ray burst data in order to reduce systematic errors on NS radius estimates (e.g., Sánchez-Fernández et al 2020).…”

“…It is worth emphasizing that it is challenging to infer the magnitude of any change in the accretion rate through spectral analysis. While Sánchez-Fernández et al (2020) found that the normalization factor for the persistent emission rises by 30-80%, other analyses show increases of 2 (e.g., in't Zand et al 2013;Worpel et al 2015;Degenaar et al 2016;Keek et al 2018). As Sánchez-Fernández et al (2020) notes, these larger increases in accretion rate could be due to only using the normalization factor to track the evolution of the persistent spectrum, or because the difference in persistent emission increase is model dependent.…”

“…The earliest example was discovered by Maccarone & Coppi (2003) who found that the 30-60 keV emission dropped by a factor of 2 during a burst from Aql X-1. Stacking of bursts have led to several measurements with large significance, most recently by Sánchez-Fernández et al (2020) who detected an 80% emission shortage in the 35-70 keV band after stacking several bursts from GS 1826-238. As these changes occur at much higher energies than found in the burst spectrum, they appear to be evidence of changes in the corona due to the burst, likely from Compton cooling due to the large influx of soft burst photons (e.g., Maccarone & Coppi 2003;Ji et al 2014a,b;Degenaar et al 2018;Chen et al 2018;Fragile et al 2018;Sánchez-Fernández et al 2020).…”

Cooling of Accretion Disc Coronae by Type I X-ray Bursts

“…Stacking of bursts have led to several measurements with large significance, most recently by Sánchez-Fernández et al (2020) who detected an 80% emission shortage in the 35-70 keV band after stacking several bursts from GS 1826-238. As these changes occur at much higher energies than found in the burst spectrum, they appear to be evidence of changes in the corona due to the burst, likely from Compton cooling due to the large influx of soft burst photons (e.g., Maccarone & Coppi 2003;Ji et al 2014a,b;Degenaar et al 2018;Chen et al 2018;Fragile et al 2018;Sánchez-Fernández et al 2020). It appears that a close investigation of the hard X-ray flux properties during a burst may therefore provide an unique probe of the corona.…”