240
D. K. Galloway and L. Keek
the fuel is hydrogen-rich, however, nuclear burning via the rp-process can prolong
the burst for up to ≈100 s [e.g., 159].
Historically the burst timescale has been measured in a number of ways that
sometimes make it difficult to compare different analyses. Measurements that rely
on the length of the interval during which the burst emission exceeds some factor
of the persistent noise level, are subject to variations due to different instrumental
sensitivities. Burst decays have been fitted with single (or sometimes multiple)
exponential decay segments, with the decay timescale quoted as a measure of the
burst duration (e.g., [53]). In high signal-to-noise lightcurves, such models generally
offer poor fits to the decay, and in fact more recent analysis favours power-law or
more complex combinations [83, 85]. A third, oft-quoted measure is the ratio of the
burst fluence to the peak flux, i.e. τ = E b /F peak (e.g., [187]). However, this quantity
does not offer an unambiguous mapping to the burst fuel composition. For example,
mixed H/He bursts with long rp-process tails may have similar τ -values as long,
intense PRE bursts arising from ignition of a deep pile of pure He.
In rare cases bursts have much longer durations, of minutes or tens of minutes
(intermediate duration bursts) to many hours (so-called “superbursts”). These
durations cannot be explained by prolonged burning, but are comparable to the
cooling timescales of deeper layers. As it takes a longer time to accumulate a larger
fuel layer, these bursts have long recurrence times of months to years. Observations
are, therefore, rare, and have mostly been performed with wide-field or all-sky
instruments, which yield data of modest quality.
The long bursts reach similar peak fluxes and photospheric temperatures as
regular bursts. The intermediate duration bursts all reach the Eddington limit,
whereas most superbursts have a lower peak flux. From an observer’s perspective,
the main discriminating characteristic is the long duration and correspondingly large
fluence of these events (Table 5.2).
Long bursts are of particular interest, because they probe deeper regions of the
neutron star envelope down to the outer crust. Furthermore, long durations allow
for more detailed spectra to be obtained, displaying interesting behaviour such as
interaction with the accretion environment (Sect. 5.4).
5.7.1 Intermediate Duration Bursts
Intermediate burst durations have been observed with exponential decay timescales
of several minutes up to ∼40 min. These events are thought to be powered by deep
helium ignition. At low mass accretion rates of ≈0.001–0.01 ˙
M Edd , the neutron
star envelope is relatively cool, allowing a large helium pile to accumulate prior
to ignition. Furthermore, at low temperatures, the ignition depth strongly increases
with temperature (Fig. 5.2), such that small temperature variations produce bursts
with a large range of ignition depths and durations. The durations can be anywhere
between a few minutes to tens of minutes, and may even rival the hours-long
superbursts (Sect. 5.7.2). In a few cases of very low persistent flux, the burst decay
D. K. Galloway and L. Keek
the fuel is hydrogen-rich, however, nuclear burning via the rp-process can prolong
the burst for up to ≈100 s [e.g., 159].
Historically the burst timescale has been measured in a number of ways that
sometimes make it difficult to compare different analyses. Measurements that rely
on the length of the interval during which the burst emission exceeds some factor
of the persistent noise level, are subject to variations due to different instrumental
sensitivities. Burst decays have been fitted with single (or sometimes multiple)
exponential decay segments, with the decay timescale quoted as a measure of the
burst duration (e.g., [53]). In high signal-to-noise lightcurves, such models generally
offer poor fits to the decay, and in fact more recent analysis favours power-law or
more complex combinations [83, 85]. A third, oft-quoted measure is the ratio of the
burst fluence to the peak flux, i.e. τ = E b /F peak (e.g., [187]). However, this quantity
does not offer an unambiguous mapping to the burst fuel composition. For example,
mixed H/He bursts with long rp-process tails may have similar τ -values as long,
intense PRE bursts arising from ignition of a deep pile of pure He.
In rare cases bursts have much longer durations, of minutes or tens of minutes
(intermediate duration bursts) to many hours (so-called “superbursts”). These
durations cannot be explained by prolonged burning, but are comparable to the
cooling timescales of deeper layers. As it takes a longer time to accumulate a larger
fuel layer, these bursts have long recurrence times of months to years. Observations
are, therefore, rare, and have mostly been performed with wide-field or all-sky
instruments, which yield data of modest quality.
The long bursts reach similar peak fluxes and photospheric temperatures as
regular bursts. The intermediate duration bursts all reach the Eddington limit,
whereas most superbursts have a lower peak flux. From an observer’s perspective,
the main discriminating characteristic is the long duration and correspondingly large
fluence of these events (Table 5.2).
Long bursts are of particular interest, because they probe deeper regions of the
neutron star envelope down to the outer crust. Furthermore, long durations allow
for more detailed spectra to be obtained, displaying interesting behaviour such as
interaction with the accretion environment (Sect. 5.4).
5.7.1 Intermediate Duration Bursts
Intermediate burst durations have been observed with exponential decay timescales
of several minutes up to ∼40 min. These events are thought to be powered by deep
helium ignition. At low mass accretion rates of ≈0.001–0.01 ˙
M Edd , the neutron
star envelope is relatively cool, allowing a large helium pile to accumulate prior
to ignition. Furthermore, at low temperatures, the ignition depth strongly increases
with temperature (Fig. 5.2), such that small temperature variations produce bursts
with a large range of ignition depths and durations. The durations can be anywhere
between a few minutes to tens of minutes, and may even rival the hours-long
superbursts (Sect. 5.7.2). In a few cases of very low persistent flux, the burst decay