Play from the GRS 1915+105 heartbeat paper
Take its pulse
In this Swift observation the black hole GRS 1915+105 beat 21 times in about 17 minutes. The trace below is that real count rate, cut into the five phases from the paper. Each phase has its own spectral fit, and those fits drive the 3D disc and corona. So you can watch one beat from the inside, and then check how much of it holds up once the hard X-rays come in.
Loading the 3D view.
Between the five phases the scene blends smoothly from one fitted value to the next.
Loading the light curve.
Swift XRT count rate, counts per second
OBSID 00030333057, 1 August 2014, 0.2 to 10 keV, 1 s bins- 1 rise, first third
- 2 rise, middle
- 3 rise, last third
- 4 fall, first half
- 5 fall, second half
What you're seeing
The trace is the Swift XRT count rate from OBSID 00030333057, taken on 1 August 2014, the observation plotted in Fig. 3 of the paper. It comes from my own reduction: Windowed Timing mode, a 40 arcsec circle around the source, everything from 0.2 to 10 keV, 1 s bins, and no background subtracted. The colours are the paper's phases. Each climb from a trough to a peak is cut into three equal pieces (phases 1 to 3) and each drop back down into two (4 and 5), and matching pieces from all the cycles were stacked before fitting.
The 3D view only knows five numbers per quantity, one per phase, because that's all the fits give. In between I blend smoothly from one phase's value to the next, so every in-between state is my interpolation. The disc colour is the fitted inner disc temperature on one fixed scale from 1.2 to 2.0 keV. The bright zone on the inner disc grows and shrinks with the apparent inner radius from the fit. The corona gets brighter with the normalisation of the Comptonisation component and bluer with its electron temperature. The whole scene also glows a little brighter when the count rate goes up.
Swift sees 1 to 10 keV. With that alone the disc swings from 1.53 keV in phase 3 to 1.91 keV in phase 4, and its apparent radius from 37.6 km down to 23.4 km (Table 2). AstroSat's two instruments cover 0.8 to 30 keV together. Fitting all five AstroSat phases at once with a single disc temperature gives 1.275 ± 0.020 keV, and it costs Δχ² = +4.3 for 4 extra constraints, so the data are fine with it. Holding the disc's apparent area fixed as well costs Δχ² = +175.9 and is rejected, so something about the disc still changes, about 20% in apparent radius. The corona carries the rest. In that tied-temperature fit its electron temperature climbs from 6.5 keV in phase 1 to 13.8 keV in phase 5 (Table 3). Fitting each phase on its own gives 6.2 keV in phase 1, rising to 14.5 keV in phase 4 (Table 2).
What this toy leaves out
The two modes are two telescopes on different days. Swift's 24 observations run from May 2014 to April 2015 (Table 1), AstroSat's two are from 14 and 15 April 2017, and the fit models differ too, so the trace is always the Swift one and in AstroSat mode it only tells the scene which phase it's in. The paper puts the bigger Swift swings down to the narrow band, where the corona can't be pinned down and the disc numbers soak up its changes. Both modes share one colour scale, so the AstroSat disc also looks cooler overall. Compare the swing inside a mode more than the level between them.
The apparent area is a fit number. It depends on the distance, the angle we see the disc at and a colour correction, so the bright zone growing here doesn't mean the inner edge of the disc moves, and in the scene the edge stays put. The paper leans towards a changing colour correction for that last 20% (as in Zoghbi et al. 2016) but doesn't settle it.
Nothing here tells you what starts each beat. The phases are bins and the source keeps changing inside each one. Sizes are not to scale, and the corona's shape is made up, since a spectral fit has no shape in it.
The paper page, in plain words
References
- Akbari, K., Patel, C., Bhattacharya, S., Bhattacharyya, S., & Choudhury, M. 2026, ApJ, 1007, 64. Probing heartbeat oscillations from the black hole X-ray binary GRS 1915+105 using spectral-timing analysis. Tables 2 and 3, Fig. 3.
- Zoghbi, A., et al. 2016, ApJ, 833, 165. Disk–wind connection during the heartbeats of GRS 1915+105.