230
D. K. Galloway and L. Keek
from 4U 0614+091 and 4U 1722−30 [76]. These detections were made with the
RXTE/PCA, which has modest spectral resolution (ΔE/E ≈ 0.17 at 6 keV, where
ΔE is the full-width at half maximum; [87]). Improved constraints on the neutron
star equation of state will result from future observations of edges in helium-rich
bursts performed with high-resolution spectrometers, but it has proven challenging
to schedule such observations of relatively infrequent helium flashes.
5.4 Interaction with the Accretion Environment
The dynamical influence of thermonuclear bursts on the accretion disk has generally
been studied in the context of radius-expansion bursts, which exceed the Eddington
limit at the neutron star surface, and may be expected to drive winds into and across
the disk. More broadly, three principal effects have been discussed in the context
of interactions between bursts and the accretion disk (e.g. [8]): structural changes
to the disk (and/or corona) resulting from burst-induced X-ray heating (or cooling);
radiatively (or thermally) driven outflows; and inflow due to Poynting-Robertson
drag. In addition, the observed spectrum may be modified by interaction with, and
reflection from, the accretion disk.
The evidence for structural changes comes from some observations of longduration bursts (see also Sect. 5.7). In some of these events we observe strong flux
variability late in the tail of the burst [37, 81] (Fig. 5.10). This variability has been
suggested to arise from Thompson scattering in the settling “fragments” of disk
material, following disruption by an extremely energetic radius-expansion burst.
Decreases in the intensity of high-energy (30–50 keV) photons beginning a few
seconds after the burst onset have also been observed in a few sources [88, 121].
These photons are thought to arise from Compton scattering in a hot corona, and it
is expected that the sudden rise in (relatively) cool photons will result in a transient
change in the coronal structure. While the principal effect predicted is additional
cooling of the corona, the estimated change in the cooling rates is in the other
direction, i.e. predicting reduced cooling [89]. Such explanations also seem at odds
with the evidence supporting the role of magnetic fields energizing coronae in other
accreting sources (e.g. [132]). Other possible mechanisms which might instead give
the observed behaviour are related to the attenuation of the inner disk in response
to the burst (see below), possibly mediated via the magnetic field threading the disk
[88].
Irradiation of the accretion disk may drive a wind from its surface. The prospects
of this phenomenon have been discussed in the context of spectral features observed
from an intermediate duration burst of IGR J17062−6143 [37]. Additionally, such
powerful bursts may eject a shell from the atmosphere of the neutron star in a
superexpansion phase, and enrich the surroundings in metal-rich burst ashes [196].
This mechanism has been argued to be the source of discrete spectral features
detected in intermediate duration bursts ([76]; see also Sect. 5.3.2).
D. K. Galloway and L. Keek
from 4U 0614+091 and 4U 1722−30 [76]. These detections were made with the
RXTE/PCA, which has modest spectral resolution (ΔE/E ≈ 0.17 at 6 keV, where
ΔE is the full-width at half maximum; [87]). Improved constraints on the neutron
star equation of state will result from future observations of edges in helium-rich
bursts performed with high-resolution spectrometers, but it has proven challenging
to schedule such observations of relatively infrequent helium flashes.
5.4 Interaction with the Accretion Environment
The dynamical influence of thermonuclear bursts on the accretion disk has generally
been studied in the context of radius-expansion bursts, which exceed the Eddington
limit at the neutron star surface, and may be expected to drive winds into and across
the disk. More broadly, three principal effects have been discussed in the context
of interactions between bursts and the accretion disk (e.g. [8]): structural changes
to the disk (and/or corona) resulting from burst-induced X-ray heating (or cooling);
radiatively (or thermally) driven outflows; and inflow due to Poynting-Robertson
drag. In addition, the observed spectrum may be modified by interaction with, and
reflection from, the accretion disk.
The evidence for structural changes comes from some observations of longduration bursts (see also Sect. 5.7). In some of these events we observe strong flux
variability late in the tail of the burst [37, 81] (Fig. 5.10). This variability has been
suggested to arise from Thompson scattering in the settling “fragments” of disk
material, following disruption by an extremely energetic radius-expansion burst.
Decreases in the intensity of high-energy (30–50 keV) photons beginning a few
seconds after the burst onset have also been observed in a few sources [88, 121].
These photons are thought to arise from Compton scattering in a hot corona, and it
is expected that the sudden rise in (relatively) cool photons will result in a transient
change in the coronal structure. While the principal effect predicted is additional
cooling of the corona, the estimated change in the cooling rates is in the other
direction, i.e. predicting reduced cooling [89]. Such explanations also seem at odds
with the evidence supporting the role of magnetic fields energizing coronae in other
accreting sources (e.g. [132]). Other possible mechanisms which might instead give
the observed behaviour are related to the attenuation of the inner disk in response
to the burst (see below), possibly mediated via the magnetic field threading the disk
[88].
Irradiation of the accretion disk may drive a wind from its surface. The prospects
of this phenomenon have been discussed in the context of spectral features observed
from an intermediate duration burst of IGR J17062−6143 [37]. Additionally, such
powerful bursts may eject a shell from the atmosphere of the neutron star in a
superexpansion phase, and enrich the surroundings in metal-rich burst ashes [196].
This mechanism has been argued to be the source of discrete spectral features
detected in intermediate duration bursts ([76]; see also Sect. 5.3.2).