220
D. K. Galloway and L. Keek
matches the expectation of atmosphere models (e.g. [178]; see also Fig. 5.4). For
this reason, these bursts have been adopted by some authors as the only reliable
events from which the neutron star mass and radius can be inferred (e.g. [91]; see
also Sect. 5.3.1).
5.1.2.2 High (Soft) State Bursts
When the persistent spectrum is soft (the “banana” state in colour-colour diagrams)
the characteristics of bursts are remarkably different. Bursts are irregular, with
longer recurrence times (on average), and exhibiting rapid (1 s) rises and short
(≈10–20 s) durations characteristic of primarily He fuel, even in those systems
which accrete mixed H/He. Radius-expansion is more frequently seen, as are burst
oscillations. The burst rises tend to have a concave shape, suggesting the ignition
location has moved away from the equator. The blackbody normalisation is typically
flat in the burst tail, which has motivated some observers to use the average as
indicative of the projected radius [64]; but as this behaviour is at odds with the
expected behaviour (arising from the flux-dependent distortion of the underlying
blackbody component) this approach has raised some objections.
We highlight this dichotomy here because the physical mechanism is not
presently understood, but clearly has a wide impact on a range of investigations
involving bursts. It is often assumed that the burst behaviour is independent of the
details of the mechanism by which the accreted fuel arrives onto the neutron star
(see Sect. 5.1.1.1); the observations seem to suggest this cannot be true. That is, the
observations suggest that the disk geometry (inferred from the persistent spectral
state) influences the burst physics, in addition to the accretion rate. Alternatively,
it is possible that the different accretion regimes give rise to different (local)
conditions, perhaps allowing some ongoing level of steady He-burning. Such effects
would certainly explain the weak bursts typically observed in the high (soft) state,
as well as the decrease in burst rate at increasing accretion rates observed for many
systems [28]. As we will see, one of the growing areas of research over the past
decade has been the interaction between the bursts and the accretion disk and
environment of the neutron star (Sect. 5.4).
5.1.2.3 The Rossi X-ray Timing Explorer
Although many X-ray instruments have contributed to our present understanding of
thermonuclear bursts, a key instrument that has enabled much of the new results
in burst phenomenology over the last decade is the Proportional Counter Array
(PCA; [86]) on-board the Rossi X-ray Timing Explorer (RXTE). This instrument,
with an effective area of ≈6500 cm 2 and fast timing down to 1 μs has allowed
detailed studies of many prolific burst sources, with high signal-to-noise data
permitting precise measurements of spectral and timing properties. The assembly of
the accumulated ≈1100 bursts over more than a decade of observations from more
D. K. Galloway and L. Keek
matches the expectation of atmosphere models (e.g. [178]; see also Fig. 5.4). For
this reason, these bursts have been adopted by some authors as the only reliable
events from which the neutron star mass and radius can be inferred (e.g. [91]; see
also Sect. 5.3.1).
5.1.2.2 High (Soft) State Bursts
When the persistent spectrum is soft (the “banana” state in colour-colour diagrams)
the characteristics of bursts are remarkably different. Bursts are irregular, with
longer recurrence times (on average), and exhibiting rapid (1 s) rises and short
(≈10–20 s) durations characteristic of primarily He fuel, even in those systems
which accrete mixed H/He. Radius-expansion is more frequently seen, as are burst
oscillations. The burst rises tend to have a concave shape, suggesting the ignition
location has moved away from the equator. The blackbody normalisation is typically
flat in the burst tail, which has motivated some observers to use the average as
indicative of the projected radius [64]; but as this behaviour is at odds with the
expected behaviour (arising from the flux-dependent distortion of the underlying
blackbody component) this approach has raised some objections.
We highlight this dichotomy here because the physical mechanism is not
presently understood, but clearly has a wide impact on a range of investigations
involving bursts. It is often assumed that the burst behaviour is independent of the
details of the mechanism by which the accreted fuel arrives onto the neutron star
(see Sect. 5.1.1.1); the observations seem to suggest this cannot be true. That is, the
observations suggest that the disk geometry (inferred from the persistent spectral
state) influences the burst physics, in addition to the accretion rate. Alternatively,
it is possible that the different accretion regimes give rise to different (local)
conditions, perhaps allowing some ongoing level of steady He-burning. Such effects
would certainly explain the weak bursts typically observed in the high (soft) state,
as well as the decrease in burst rate at increasing accretion rates observed for many
systems [28]. As we will see, one of the growing areas of research over the past
decade has been the interaction between the bursts and the accretion disk and
environment of the neutron star (Sect. 5.4).
5.1.2.3 The Rossi X-ray Timing Explorer
Although many X-ray instruments have contributed to our present understanding of
thermonuclear bursts, a key instrument that has enabled much of the new results
in burst phenomenology over the last decade is the Proportional Counter Array
(PCA; [86]) on-board the Rossi X-ray Timing Explorer (RXTE). This instrument,
with an effective area of ≈6500 cm 2 and fast timing down to 1 μs has allowed
detailed studies of many prolific burst sources, with high signal-to-noise data
permitting precise measurements of spectral and timing properties. The assembly of
the accumulated ≈1100 bursts over more than a decade of observations from more