A number of neutron star low-mass X-ray binaries (LMXBs) have recently been discovered to show broad, asymmetric Fe K emission lines in their X-ray spectra. These lines are generally thought to be ...the most prominent part of a reflection spectrum, originating in the inner part of the accretion disk where strong relativistic effects can broaden emission lines. We present a comprehensive, systematic analysis of Suzaku and XMM-Newton spectra of 10 neutron star LMXBs, all of which display broad Fe K emission lines. Of the 10 sources, 4 are Z sources, 4 are atolls, and 2 are accreting millisecond X-ray pulsars (also atolls). The Fe K lines are fit well by a relativistic line model for a Schwarzschild metric, and imply a narrow range of inner disk radii (6-15 GM/c {sup 2}) in most cases. This implies that the accretion disk extends close to the neutron star surface over a range of luminosities. Continuum modeling shows that for the majority of observations, a blackbody component (plausibly associated with the boundary layer) dominates the X-ray emission from 8 to 20 keV. Thus it appears likely that this spectral component produces the majority of the ionizing flux that illuminates the accretion disk. Therefore, we also fit the spectra with a blurred reflection model, wherein a blackbody component illuminates the disk. This model fits well in most cases, supporting the idea that the boundary layer illuminates a geometrically thin disk.
With an inferred bolometric luminosity exceeding 1042 erg s--1, HLX-1 in ESO 243-49 is the most luminous of ultraluminous X-ray sources and provides one of the strongest cases for the existence of ...intermediate-mass black holes. We obtain good fits to disk-dominated observations of the source with BHSPEC, a fully relativistic black hole accretion disk spectral model. Due to degeneracies in the model arising from the lack of independent constraints on inclination and black hole spin, there is a factor of 100 uncertainty in the best-fit black hole mass M. Nevertheless, spectral fitting of XMM-Newton observations provides robust lower and upper limits with 3000 M M 3 X 105 M , at 90% confidence, placing HLX-1 firmly in the intermediate-mass regime. The lower bound on M is entirely determined by matching the shape and peak energy of the thermal component in the spectrum. This bound is consistent with (but independent of) arguments based solely on the Eddington limit. Joint spectral modeling of the XMM-Newton data with more luminous Swift and Chandra observations increases the lower bound to 6000 M , but this tighter constraint is not independent of the Eddington limit. The upper bound on M is sensitive to the maximum allowed inclination i, and is reduced to M 105 M if we limit i 75?.
Global warming imposes us to reflect on the way we carry research, embarking on the obligation to minimize the environmental impact of our research programs, with the reduction of our travel ...footprint being one of the easiest actions to implement, thanks to the advance of digital technology. The X-ray Integral Field Unit (X-IFU), the cryogenic spectrometer of the Athena space X-ray observatory of the European Space Agency will be developed by a large international consortium, currently involving 240 members, split over 13 countries, 11 in Europe, Japan and the United States. The travel footprint associated with the development of the X-IFU is to be minimized. For that purpose, a travel footprint calculator has been developed and released to the X-IFU consortium members. The calculator uses seven different emission factors and methods leading to estimates that differ by up to a factor of 5 for the same flying distance. These differences illustrate the lack of standards and regulations for computing the footprint of flight travels and are explained primarily, though partly, by different accounting of non- CO2 effects. When accounting for non-CO2 effects, the flight emission is estimated as a multiple of the direct CO2 emission from burning fuel, expressed in CO2-equivalent (CO2eq), with a multiplication factor ranging from 2 to 3. Considering or ignoring this multiplication factor is key when comparing alternative modes of transportation to flying. The calculator enables us to compute the travel footprint of a large set of travels and can help identify a meeting place that minimizes the overall travel footprint for a large set of possible city hosts, e.g. cities with large airports. The calculator also includes the option for a minimum distance above which flying is considered the most suitable transport option; below that chosen distance, the emission of train journeys are considered. To demonstrate its full capabilities, the calculator is first run on one of the largest scientific meetings; the fall meeting of the American Geoscience Union (AGU) gathering some 24000 participants and the four meetings of the lead authors of the working group I of the Intergovernmental Panel on Climate Change (IPCC) preparing its sixth assessment report. In both examples, the calculator is used to compute the location of the meetings that would minimize the travel footprint. Then, the travel footprint of a representative set of X-IFU related meetings is estimated to be 500 tons of CO2eq per year (to place this number in perspective, it is equivalent to 2 billion kilometers driven by an average passenger vehicle). Of this amount, each annual consortium meeting accounts for 100 tons, being located at a site of minimum emission and for a minimum distance for flying of 700 km. Actions to reduce the X-IFU travel footprint are being implemented, e.g., the number of large consortium meetings has been reduced to one per year and face-to-face working meetings are progressively replaced by video conferences. As the on-line travel footprint calculator may be used to all scientific collaborations and meetings, the calculator and its methodology described in this paper are made freely available to the science communitycommunity(
https://travel-footprint-calculator.irap.omp.eu
).
Z sources are bright neutron star X-ray binaries, accreting at around the Eddington limit. We analyze the 68 RXTE observations (~270 ks) of Sco-like Z source GX 17+2 made between 1999 October 3 and ...12, covering a complete Z track. We create and fit color-resolved spectra with a model consisting of a thermal multicolor disk, a single-temperature-blackbody boundary layer and a weak Comptonized component. We find that, similar to what was observed for XTE J1701-462 in its Sco-like Z phase, the branches of GX 17+2 can be explained by three processes operating at a constant accretion rate M into the disk: increase of Comptonization up the horizontal branch (HB), transition from a standard thin disk to a slim disk up the normal branch (NB), and temporary fast decrease of the inner disk radius up the flaring branch. We also model the Comptonization in an empirically self-consistent way, with its seed photons tied to the thermal disk component and corrected for to recover the pre-Comptonized thermal disk emission. This allows us to show a constant M along the entire Z track based on the thermal disk component. We also measure the upper kHz quasi-periodic oscillation frequency and find it to depend on the apparent inner disk radius R sub(in) (prior to Compton scattering) approximately as frequency infinity (ProQuest: Formulae and/or non-USASCII text omitted), supporting the identification of it as the Keplerian frequency at R sub(in). The HB oscillation is probably related to the dynamics in the inner disk as well, as both its frequency and R sub(in) vary significantly on the HB but become relatively constant on the NB.
Orbital Decay in M82 X-2 Bachetti, Matteo; Heida, Marianne; Maccarone, Thomas ...
The Astrophysical journal,
10/2022, Letnik:
937, Številka:
2
Journal Article
Recenzirano
Odprti dostop
Abstract
M82 X-2 is the first pulsating ultraluminous X-ray source discovered. The luminosity of these extreme pulsars, if isotropic, implies an extreme mass transfer rate. An alternative is to ...assume a much lower mass transfer rate, but with an apparent luminosity boosted by geometrical beaming. Only an independent measurement of the mass transfer rate can help discriminate between these two scenarios. In this paper, we follow the orbit of the neutron star for 7 yr, measure the decay of the orbit (
P
̇
orb
/
P
orb
≈
−
8
·
10
−
6
yr
−
1
), and argue that this orbital decay is driven by extreme mass transfer of more than 150 times the mass transfer limit set by the Eddington luminosity. If this is true, the mass available to the accretor is more than enough to justify its luminosity, with no need for beaming. This also strongly favors models where the accretor is a highly magnetized neutron star.
Context. Neural networks are being extensively used for modeling data, especially in the case where no likelihood can be formulated. Aims. Although in the case of X-ray spectral fitting the ...likelihood is known, we aim to investigate the ability of neural networks to recover the model parameters and their associated uncertainties and to compare their performances with standard X-ray spectral fitting, whether following a frequentist or Bayesian approach. Methods. We applied a simulation-based inference with neural posterior estimation (SBI-NPE) to X-ray spectra. We trained a network with simulated spectra generated from a multiparameter source emission model folded through an instrument response, so that it learns the mapping between the simulated spectra and their parameters and returns the posterior distribution. The model parameters are sampled from a predefined prior distribution. To maximize the efficiency of the training of the neural network, while limiting the size of the training sample to speed up the inference, we introduce a way to reduce the range of the priors, either through a classifier or a coarse and quick inference of one or multiple observations. For the sake of demonstrating working principles, we applied the technique to data generated from and recorded by the NICER X-ray instrument, which is a medium-resolution X-ray spectrometer covering the 0.2–12 keV band. We consider here simple X-ray emission models with up to five parameters. Results. SBI-NPE is demonstrated to work equally well as standard X-ray spectral fitting, both in the Gaussian and Poisson regimes, on simulated and real data, yielding fully consistent results in terms of best-fit parameters and posterior distributions. The inference time is comparable to or smaller than the one needed for Bayesian inference when involving the computation of large Markov chain Monte Carlo chains to derive the posterior distributions. On the other hand, once properly trained, an amortized SBI-NPE network generates the posterior distributions in no time (less than 1 second per spectrum on a 6-core laptop). We show that SBI-NPE is less sensitive to local minima trapping than standard fit statistic minimization techniques. With a simple model, we find that the neural network can be trained equally well on dimension-reduced spectra via a principal component decomposition, leading to a faster inference time with no significant degradation of the posteriors. Conclusions. We show that simulation-based inference with neural posterior estimation is a complementary tool for X-ray spectral analysis. The technique is robust and produces well-calibrated posterior distributions. It holds great potential for its integration in pipelines developed for processing large data sets. The code developed to demonstrate the first working principles of the technique introduced here is released through a Python package called SIXSA (Simulation-based Inference for X-ray Spectral Analysis), which is available from GitHub.
Context. Neural networks are being extensively used for modeling data, especially in the case where no likelihood can be formulated. Aims. Although in the case of X-ray spectral fitting the ...likelihood is known, we aim to investigate the ability of neural networks to recover the model parameters and their associated uncertainties and to compare their performances with standard X-ray spectral fitting, whether following a frequentist or Bayesian approach. Methods. We applied a simulation-based inference with neural posterior estimation (SBI-NPE) to X-ray spectra. We trained a network with simulated spectra generated from a multiparameter source emission model folded through an instrument response, so that it learns the mapping between the simulated spectra and their parameters and returns the posterior distribution. The model parameters are sampled from a predefined prior distribution. To maximize the efficiency of the training of the neural network, while limiting the size of the training sample to speed up the inference, we introduce a way to reduce the range of the priors, either through a classifier or a coarse and quick inference of one or multiple observations. For the sake of demonstrating working principles, we applied the technique to data generated from and recorded by the NICER X-ray instrument, which is a medium-resolution X-ray spectrometer covering the 0.2–12 keV band. We consider here simple X-ray emission models with up to five parameters. Results. SBI-NPE is demonstrated to work equally well as standard X-ray spectral fitting, both in the Gaussian and Poisson regimes, on simulated and real data, yielding fully consistent results in terms of best-fit parameters and posterior distributions. The inference time is comparable to or smaller than the one needed for Bayesian inference when involving the computation of large Markov chain Monte Carlo chains to derive the posterior distributions. On the other hand, once properly trained, an amortized SBI-NPE network generates the posterior distributions in no time (less than 1 second per spectrum on a 6-core laptop). We show that SBI-NPE is less sensitive to local minima trapping than standard fit statistic minimization techniques. With a simple model, we find that the neural network can be trained equally well on dimension-reduced spectra via a principal component decomposition, leading to a faster inference time with no significant degradation of the posteriors. Conclusions. We show that simulation-based inference with neural posterior estimation is a complementary tool for X-ray spectral analysis. The technique is robust and produces well-calibrated posterior distributions. It holds great potential for its integration in pipelines developed for processing large data sets. The code developed to demonstrate the first working principles of the technique introduced here is released through a Python package called SIXSA (Simulation-based Inference for X-ray Spectral Analysis), which is available from GitHub.
Context. Active galactic nuclei (AGNs) display complex X-ray spectra that exhibit a variety of emission and absorption features. These are commonly interpreted as a combination of (i) a ...relativistically smeared reflection component, resulting from the irradiation of an accretion disk by a compact hard X-ray source; (ii) one or several warm or ionized absorption components produced by AGN-driven outflows crossing our line of sight; and (iii) a nonrelativistic reflection component produced by more distant material. Disentangling these components via detailed model fitting could be used to constrain the black hole spin, geometry, and characteristics of the accretion flow, as well as of the outflows and surroundings of the black hole. Aims. We investigate how a high-throughput high-resolution X-ray spectrometer such as the Athena X-ray Integral Field Unit (X-IFU) can be used to this aim, using the state-of-the-art reflection model relxill in a lamp-post geometrical configuration. Methods. We simulated a representative sample of AGN spectra, including all necessary model complexities, as well as a range of model parameters going from standard to more extreme values, and considered X-ray fluxes that are representative of known AGN and quasar populations. We also present a method to estimate the systematic errors related to the uncertainties in the calibration of the X-IFU. Results. In a conservative setting, in which the reflection component is computed self consistently by the relxill model from the pre-set geometry and no iron overabundance, the mean errors on the spin and height of the irradiating source are < 0.05 and ∼0.2 Rg (in units of gravitational radius). Similarly, the absorber parameters (column density, ionization parameter, covering factor, and velocity) are measured to an accuracy typically less than ∼5% over their allowed range of variations. Extending the simulations to include blueshifted ultra-fast outflows, we show that X-IFU could measure their velocity with statistical errors < 1%, even for high-redshift objects (e.g., at redshifts ∼2.5). Conclusion. The simulations presented here demonstrate the potential of the X-IFU to understand how black holes are powered and how they shape their host galaxies. The accuracy in recovering the physical model parameters encoded in their X-ray emission is reached thanks to the unique capability of X-IFU to separate and constrain narrow and broad emission and absorption components.
We propose the development of X-ray interferometry (XRI), to reveal the Universe at high energies with ultra-high spatial resolution. With baselines which can be accommodated on a single spacecraft, ...XRI can reach 100 μ as resolution at 10 Å (1.2 keV) and 20 μ as at 2 Å (6 keV), enabling imaging and imaging-spectroscopy of (for example) X-ray coronae of nearby accreting supermassive black holes (SMBH) and the SMBH ‘shadow’; SMBH accretion flows and outflows; X-ray binary winds and orbits; stellar coronae within ∼100 pc and many exoplanets which transit across them. For sufficiently luminous sources XRI will resolve sub-pc scales across the entire observable Universe, revealing accreting binary SMBHs and enabling trigonometric measurements of the Hubble constant with X-ray light echoes from quasars or explosive transients. A multi-spacecraft ‘constellation’ interferometer would resolve well below 1 μ as, enabling SMBH event horizons to be resolved in many active galaxies and the detailed study of the effects of strong field gravity on the dynamics and emission from accreting gas close to the black hole.
PDF
