6 High-Frequency Variability in Neutron-Star Low-Mass X-ray Binaries
265
For Neutron-Star Low-Mass X-ray Binaries (NS LMXB), the first important
result was the discovery of Quasi-Periodic Oscillations (QPO) in GX 5-1 in 1985
with the EXOSAT satellite [162], obtained serendipitously when looking for Xray pulsations (see the chapter by Patruno and Watts, this volume). From this first
detection, QPOs were soon observed in many other sources of the same class. The
detection of a signal with a very defined frequency provided the first precise time
scale measurement of the accretion flow. The first interpretation of QPOs was in
terms of the spin frequency of the neutron star. The values of their frequencies
(typically ∼1 Hz) and the fact that they were not coherent or constant in time
excluded that they could be a direct observation of the spin. However, models that
interpreted them as a beat between the rotation of the neutron star and the orbital
motion at the inner radii of the accretion flow were proposed and were able to
explain the observations (see e.g. [6, 88]). For this to be the case, the presence of a
non-negligible magnetic field is necessary. Additional types of QPOs with different
properties were also found, the origin of which was even more difficult to explain.
A few years later, similar QPOs started being observed from black-hole binaries
(BHB), thanks to the new all-sky monitors that were able to discover transient
systems (since most of the BHB are transient). Their frequency was lower (1–
10 Hz), but they appeared to be rather similar to those in NS LMXB (see e.g. [118]).
In addition, the broad-band noise component connected to QPOs was found, at least
in some source states, to be extremely similar between the two classes. Since black
holes do not have a solid surface nor a magnetic field, the NS models that depend
on either could not be applied to black holes.
The launch of RXTE at the end of 1995 opened the way to the detection of highfrequency (>100 Hz) features in the PDS of accreting binaries. For NS systems,
new quasi-periodic peaks at frequencies of hundreds of Hertz, called kilohertz
QPOs (kHz QPOs), were discovered first in the brightest source, Sco X-1 [164],
then soon in many other NS LMXBs. The model involving a beat with the neutron
star spin was adapted to interpret these high frequencies, as the peaks often appear
in pairs with roughly the same separation (e.g. [50, 151]), but new data presented
problems for the model, which had to be abandoned. RXTE also allowed to bring
the BHB QPOs into a phenomenological scheme that appears to be connected to
that of NS LMXB low-frequency QPOs [34]. The kHz phenomenon appears to be
very common in bright NS LMXBs. RXTE also discovered high-frequency QPOs
(HFQPOs) from BHBs, but they are extremely rare to the extent that, excluding
one peculiar source that had many detections [17, 119], only a handful of them
were found in the 16 years of operation of the satellite [22]. RXTE also led to the
discovery of other fast-timing phenomena from NS LMXBs that have completely
changed our knowledge of these systems: burst oscillations, accreting millisecond
pulsars and intermittent pulsars, all of which are dealt with in other chapters of this
book.
After the end of the RXTE mission it has become much more difficult to detect
fast-timing aperiodic phenomena, because missions like XMM-Newton, Chandra
or Swift are not optimised for timing studies and do not yield the high count rates
that are needed. In the recent years, the launch of the Indian satellite Astrosat, which
265
For Neutron-Star Low-Mass X-ray Binaries (NS LMXB), the first important
result was the discovery of Quasi-Periodic Oscillations (QPO) in GX 5-1 in 1985
with the EXOSAT satellite [162], obtained serendipitously when looking for Xray pulsations (see the chapter by Patruno and Watts, this volume). From this first
detection, QPOs were soon observed in many other sources of the same class. The
detection of a signal with a very defined frequency provided the first precise time
scale measurement of the accretion flow. The first interpretation of QPOs was in
terms of the spin frequency of the neutron star. The values of their frequencies
(typically ∼1 Hz) and the fact that they were not coherent or constant in time
excluded that they could be a direct observation of the spin. However, models that
interpreted them as a beat between the rotation of the neutron star and the orbital
motion at the inner radii of the accretion flow were proposed and were able to
explain the observations (see e.g. [6, 88]). For this to be the case, the presence of a
non-negligible magnetic field is necessary. Additional types of QPOs with different
properties were also found, the origin of which was even more difficult to explain.
A few years later, similar QPOs started being observed from black-hole binaries
(BHB), thanks to the new all-sky monitors that were able to discover transient
systems (since most of the BHB are transient). Their frequency was lower (1–
10 Hz), but they appeared to be rather similar to those in NS LMXB (see e.g. [118]).
In addition, the broad-band noise component connected to QPOs was found, at least
in some source states, to be extremely similar between the two classes. Since black
holes do not have a solid surface nor a magnetic field, the NS models that depend
on either could not be applied to black holes.
The launch of RXTE at the end of 1995 opened the way to the detection of highfrequency (>100 Hz) features in the PDS of accreting binaries. For NS systems,
new quasi-periodic peaks at frequencies of hundreds of Hertz, called kilohertz
QPOs (kHz QPOs), were discovered first in the brightest source, Sco X-1 [164],
then soon in many other NS LMXBs. The model involving a beat with the neutron
star spin was adapted to interpret these high frequencies, as the peaks often appear
in pairs with roughly the same separation (e.g. [50, 151]), but new data presented
problems for the model, which had to be abandoned. RXTE also allowed to bring
the BHB QPOs into a phenomenological scheme that appears to be connected to
that of NS LMXB low-frequency QPOs [34]. The kHz phenomenon appears to be
very common in bright NS LMXBs. RXTE also discovered high-frequency QPOs
(HFQPOs) from BHBs, but they are extremely rare to the extent that, excluding
one peculiar source that had many detections [17, 119], only a handful of them
were found in the 16 years of operation of the satellite [22]. RXTE also led to the
discovery of other fast-timing phenomena from NS LMXBs that have completely
changed our knowledge of these systems: burst oscillations, accreting millisecond
pulsars and intermittent pulsars, all of which are dealt with in other chapters of this
book.
After the end of the RXTE mission it has become much more difficult to detect
fast-timing aperiodic phenomena, because missions like XMM-Newton, Chandra
or Swift are not optimised for timing studies and do not yield the high count rates
that are needed. In the recent years, the launch of the Indian satellite Astrosat, which