Search

We study orbital modulation of X-rays from Cyg X-3, using data from Swift, INTEGRAL and RXTE. Using the wealth of the presently available data and an improved averaging method, we obtain energy-dependent folded and averaged light curves with unprecedented accuracy. We find that above ∼5 keV, the modulation depth decreases with the increasing energy, which is consistent with the modulation being caused by both bound-free absorption and Compton scattering in the stellar wind of the donor, with minima corresponding to the highest optical depth, which occurs around the superior conjunction. We find a decrease of the depth below ∼3 keV, which appears to be due to re-emission of the absorbed continuum by the wind in soft X-ray lines. Based on the shape of the folded light curves, any X-ray contribution from the jet in Cyg X-3, which emits γ-rays detected at energies > 0.1 GeV in soft spectral states, is found to be minor up to ∼100 keV. This implies the presence of a rather sharp low-energy break in the jet MeV-range spectrum.We also calculate phase-resolved RXTE X-ray spectra, and show the difference between the spectra corresponding to phases around the superior and inferior conjunctions can indeed be accounted for by a combined effect of bound-free absorption in an ionized medium and Compton scattering.

The microquasar GRS 1758-258 is one of only three persistent black hole binaries that spend most of their time in the hard spectral state, the other two being Cyg X-l and 1E 1741.7-2942. It therefore provides the rare opportunity for an extensive long term study of this important black hole state which is associated with strong variability and radio jet emission. INTEGRAL has been monitoring the source since the first Galactic Center Deep Exposure season in spring 2003 during two 2-3 months long Galactic Center viewing epochs each year, amounting to 11 epochs including spring of 2008. With the exception of the last epoch quasi-simultaneous RXTE monitoring observations are available as well. Here we present an analysis of the epoch averaged broad band spectra which display considerable long term variability, most notably the occurrence of two soft/off states, extreme examples for the hysteretic behavior of black hole binaries. The hard source spectrum and long exposures allow us to extend the analysis for several epochs to approximately 800 keV using PICsIT data and address the question of the presence of a non-thermal Comptonization component.

We present a comprehensive analysis of all observations of Cyg X-1 by the Compton Gamma Ray Observatory Burst and Transient Source Experiment (BATSE; 20-300 keV) and by the Rossi X-Ray Timing Explorer all-sky monitor (ASM; 1.5-12 keV) until 2002 June, including approximately 1200 days of simultaneous data. We find a number of correlations between fluxes and hardnesses in different energy bands. In the hard (low) spectral state, there is a negative correlation between the ASM 1.5-12 keV flux and the hardness at any energy. In the soft (high) spectral state, the ASM flux is positively correlated with the ASM hardness but uncorrelated with the BATSE hardness. In both spectral states, the BATSE hardness correlates with the flux above 100 keV, while it shows no correlation with the 20-100 keV flux. At the same time, there is clear correlation between the BATSE fluxes below and above 100 keV. In the hard state, most of the variability can be explained by softening the overall spectrum with a pivot at approximately 50 keV. There is also another, independent variability pattern of lower amplitude where the spectral shape does not change when the luminosity changes. In the soft state, the variability is mostly caused by a variable hard (Comptonized) spectral component of a constant shape superposed on a constant soft blackbody component. These variability patterns are in agreement with the dependencies of the rms variability on the photon energy in the two states. We also study in detail recent soft states from late 2000 until 2002. The last of them has lasted thus far for more than 200 days. Their spectra are generally harder in the 1.5-5 keV band and similar or softer in the 3-12 keV band than the spectra of the 1996 soft state, whereas the rms variability is stronger in all the ASM bands. On the other hand, the 1994 soft state transition observed by BATSE appears very similar to the 1996 one. We interpret the variability patterns in terms of theoretical Comptonization models. In the hard state, the variability appears to be driven mostly by changing flux in seed photons Comptonized in a hot thermal plasma cloud with an approximately constant power supply. In the soft state, the variability is consistent with flares of hybrid, thermal/nonthermal, plasma with variable power above a stable cold disk. The spectral and timing differences between the 1996 and 2000-2002 soft states are explained by a decrease of the color disk temperature. Also, on the basis of broadband pointed observations simultaneous with those of the ASM and BATSE, we find the intrinsic bolometric luminosity increases by a factor of approximately 3-4 from the hard state to the soft one, which supports models of the state transition based on a change of the accretion rate.

We present a spectral analysis of observations of Cygnus X-1 by the RXTE/ASM (1.5-12 keV) and CGRO/BATSE (20-300 keV), including about 1200 days of simultaneous data. We find a number of correlations between intensities and hardnesses in different energy bands from 1.5 keV to 300 keV. In the hard (low) spectral state, there is a negative correlation between the ASM 1.5-12 keV flux and the hardness at any energy. In the soft (high) spectral state, the ASM flux is positively correlated with the ASM hardness (as previously reported) but uncorrelated with the BATSE hardness. In both spectral states, the BATSE hardness correlates with the flux above 100 keV, while it shows no correlation with the flux in the 20-100 keV range. At the same time, there is clear correlation between the BATSE fluxes below and above 100 keV. In the hard state, most of the variability can be explained by softening the overall spectrum with a pivot at approximately 50 keV. The observations show that there has to be another, independent variability pattern of lower amplitude where the spectral shape does not change when the luminosity changes. In the soft state, the variability is mostly caused by a variable hard (Comptonized) spectral component of a constant shape superimposed on a constant soft blackbody component. These variability patterns are in agreement with the dependence of the rms variability on the photon energy in the two states. We interpret the observed correlations in terms of theoretical Comptonization models. In the hard state, the variability appears to be driven mostly by changing flux in seed photons Comptonized in a hot thermal plasma cloud with an approximately constant power supply. In the soft state, the variability is consistent with flares of hybrid, thermal/nonthermal, plasma with variable power above a stable cold disk. Also, based on broadband pointed observations simultaneous with those of the ASM and BATSE, we find the intrinsic bolometric luminosity increases by a factor of approximately 3-4 from the hard state to the soft one, which supports models of the state transition based on a change of the accretion rate.

Observations and theoretical models of X-ray/gamma ray spectra of radio quiet Seyfert galaxies and Galactic black hole candidates are reviewed. The spectra from these objects share the following characteristics: an underlying power law with a high energy cutoff above 200 keV; a Compton reflection component with a Fe K alpha line, and a low energy absorption by intervening cold matter. The X-ray energy spectral index, alpha, is typically in the range between 0.8 and 1 in Seyfert spectra, and that of the hard state spectra of the black hole candidates Cygnus X-1 and GX 339-4 is typically between 0.6 and 0.8. The Compton reflection component corresponds with cold matter covering a solid angle of between 0.8pi and 2pi as seen from the X-ray source. The broadband spectra of both classes of sources are well fitted by Compton upscattering of soft photons in thermal plasma. The fits yield a thermal plasma temperature of 100 keV and the Thomson optical depth of 1. All the spectra presented are cut off before the electron rest energy 511 keV, indicating that electron/positron pair production is an important process.

We have obtained the first average 2-500 keV spectra of Seyfert galaxies, using the data from Ginga and Compton Gamma-Ray Observatory's (CGRO) Oriented Scintillation Spectrometer Experiment (OSSE). Our sample contains three classes of objects with markedly different spectra: radio-quiet Seyfert 1's and 2's, and radio-loud Seyfert 1's. The average radio-quiet Seyfert 1 spectrum is well-fitted by a power law continuum with the energy spectral index alpha approximately equals 0.9, a Compton reflection component corresponding to a approximately 2 pi covering solid angle, and ionized absorption. There is a high-energy cutoff in the incident power law continuum: the e-folding energy is E(sub c) approximately equals 0.6(sup +0.8 sub -0.3) MeV. The simplest model that describes this spectrum is Comptonization in a relativistic optically-thin thermal corona above the surface of an accretion disk. Radio-quiet Seyfert 2's show strong netural absorption, and there is an indication that their X-ray power laws are intrinsically harder. Finally, the radio-loud Seyfert spectrum has alpha approximately equals 0.7, moderate neutral absorption E(sub C) = 0.4(sup +0.7 sub -0.2) MeV, and no or little Compton reflection. This is incompatible with the radio-quiet Seyfert 1 spectrum, and probably indicating that the X-rays are beamed away from the accretion disk in these objects. The average spectra of Seyferts integrated over redshift with a power-law evolution can explain the hard X-ray spectrum of the cosmic background.

We calculate the spectrum of X-ray radiation and reprocessed by a partly ionized optically thick medium in an active galactic nucleus. We self-consistently calculate the ionization balance and thermal balance in the medium along with the distribution of X-ray intensity with optical depth. In addition to absorption or scattering of the incident X-rays, we also compute the spectrum of X-rays emitted by the material, including lines, edges, and bremsstrahlung. The albedo of the medium depends primarily on the X-ray ionization parameter (ratio of incident flux to gas density, zeta(sub Chi), and secondarily on the UV flux generated by dissipation inside the disk; we locate the critical range of zeta(sub Chi) over which the albedo increases from small to nearly unity. While the continuum reflection is very weak below 10 keV when zeta(sub Chi) is small, significnat fluxes are emitted in atomic lines and edges in this energy range. In the limit of large zeta(sub Chi), the albedo below 10 keV increases, but reflection in this band is never gray: some photoelectric absorption remains up to rather large values of zeta(sub Chi), while at still higher values, inverse Compton scattering amplifies the soft X-ray flux. These features are sufficiently sharp that current and near-future X-ray experiments should permit diagnostic measures of zeta(sub Chi).

Observations suggest the existence of both hot and cold dark matter in the centers of active galactic nuclei. Recent spectral models require a major fraction of power to be dissipated in the hot matter. We study the case when the hot matter forms a corona around a standard cold alpha-disk. In particular, we investigate the case when a major fraction, f, of the power released when the cold matter accretes is transported to and dissipated in the corona. This has major effects on the cold disk, making it colder, more geometrically thin, denser, and having larger optical depths. One important consequence is the disappearance of the effectively optically thin zone as well as of the radiation pressure dominated zone for values of f sufficiently closed to unity. The disappearance of the radiation pressure dominated zone will result in a cold disk with only a gas pressure dominated zone that is stable against thermal and viscous instabilities. We also show that the pressure ( and the radiation) from the corona will only affect the surface layers of the cold disk. Our results disagree with those of other recent work on accretion disks with coronae. We find those works to be based on unphysical assumptions.

We study possible models of two Galactic sources of transient pair annihilation radiation, 1E 1740.7-2942 and a source observed by High Energy Astronomy Observatory (HEAO) 1 A-4. We fit the observed spectral features by thermal annihilation spectra and find that the redshifts obtained by us are much larger than those obtained from fitting Caussian lines centered on 511 keV. This effect, which is due to the net blueshift (with respect to 511 keV) of the annihilation spectrum due to the thermal energies of pairs, puts strong constraints on models of sources. We consider those constraints first without considering the mechanism of positron production. From the shape of the observed spectra, we are able to rule out both spherical clouds and layers above cold matter as possible source geometries. The observed spectra are compatible with two source geometries: (1) a nearly face-on disk in the Kerr metric and (2) a jet close to a black hole. We consider, then, the origin of the pairs. Theories of both thermal and nonthermal pair equilibria predict that photon-pair production is unable to produce annihilation features that contain as much as half of the bolometric luminosity, which is observed. A possible solution to this problem is obscuration of a nonthermal source (in which pairs are produced by photon-photon collisions) and an outflow of pairs to an unobscured region. This makes annihilation in a jet the most likely model of the considered sources.

We show the effects of thermal cyclosynchrotron radiation and magnetic viscosity on the structure of hot, two-temperature accretion disks. Magnetic field, B, is assumed to be randomly oriented and the ratio of magnetic pressure to either gas pressure, alpha = P(sub mag)/P(sub gas), or the sum of the gas and radiation pressures, alpha = (P(sub mag)/P(sub gas) + P(sub rad)), is fixed. We find those effects do not change the qualitative properties of the disks, i.e., there are still two critical accretion rates related to production of e(sup +/-) pairs, (M dot)((sup U)(sub cr)) and (M dot)((sup L)(sub cr)), that affect the number of local and global disk solutions, as recently found by Bjoernsson and Svensson for the case with B = 0. However, a critical value of the alpha-viscosity parameter above which those critical accretion rates disappear becomes smaller than alpha(sub cr) = 1 found in the case of B = 0, for P(sub mag) = alpha(P(sub gas) + P(sub rad)). If P(sub mag) = alpha P(sub gas), on the other hand, alpha(sub cr) is still about unity. Moreover, when Comptonized cyclosynchrotron radiation dominates Comptonized bremsstrahlung, radiation from the disk obeys a power law with the energy spectral index of approximately 0.5, in a qualitative agreement with X-ray observations of active galactic nuclei (AGNS) and Galactic black hole candidates. We also extend the hot disk solutions for P(sub mag) = alpha(P(sub gas) + P(sub rad)) to the effectively optically thick region, where they merge with the standard cold disk solutions. We find that the mapping method by Bjoernsson and Svensson gives a good approximation to the disk structure in the hot region and show where it breaks in the transition region. Finally, we find a region in the disk parameter space with no solutions due to the inability of Coulomb heating to supply enough energy to electrons.

We show that the recent observations of the Seyfert galaxy NGC 4151 in hard X-rays and soft gamma rays by the OSSE and SIGMA detectors on board CGRO and GRANAT, respectively, are well explained by a nonthermal model with acceleration of relativistic electrons at an efficiency of less than 50 percent and with the remaining power dissipated thermally in the source (the standard nonthermal e(+/-) pair model assumed 100 percent efficiency). Such an acceleration efficiency is generally expected on physical grounds. The resulting model unifies previously proposed purely thermal and purely nonthermal models. The pure nonthermal model for NGC 4151 appears to be ruled out. The pure thermal model gives a worse fit to the data than our hybrid nonthermal/thermal model.

While it has long been known that active galactic nuclei (AGNs) make up a significant part of the X-ray background near 1 keV, it has been generally thought that the typical AGN X-ray spectrum did not fit the higher energy background spectrum. More recent data (from Ginga, SIGMA, and OSSE) show that the X-ray spectra of radio-quiet AGNs do in fact resemble the X-ray background spectrum. These data are well described (from 2 to 100 keV) by a model in which the X-ray photons are made by thermal Comptonization, and a fraction (around a half) of the X-rays are reflected by a cool surface. Using the observed evolution of X-ray-bright AGNs, we show that the known AGN population, radiating the spectrum observed at low redshift and integrated in redshift back to z about 4, is capable of producing the entire X-ray background from 2 to 100 keV.

The nearest BL Lac object, Mrk 421, has a gamma-ray spectrum which is approximately flat in EF-sub E from E less than about 50 MeV to E greater than about 1 TeV. Inverse Compton scattering can explain this smooth spectrum, despite the structure in the Klein-Nishina cross section, if the injected electron distribution function is proportional to gamma exp -2, where gamma is the electron Lorentz factor. When this is the case, the structure imprinted on the steady state electron distribution function by the structure in the Klein-Nishina cross section is almost exactly compensated in the radiated spectrum. Because particle acceleration in strong shocks injects particles with this distribution function, this shape injection function is in fact quite plausible. Other blazars may be explained by the same model if the cutoff below TeV energies observed in other objects is due to pair production on the cosmological IR background, as suggested by Stecker et al. (1992).

Consideration is given to recent models for the cosmic X-ray background that assume that it originates from unresolved AGN emitting spectra due to enhanced Compton reflection of a power-law photon spectrum incident on cold matter. The parameter space of the Compton reflection model is studied, and the allowed parameter space is found to be severely constrained by physical and cosmological effects. For an incident power-law energy index alpha is greater than about 1, the X-ray peak in the observed spectrum from a population of AGN is necessarily at an energy less than that of the observed peak. Two examples of improved fits to the X-ray background are shown. It is concluded that the Compton reflection models proposed to date do not provide a straightforward explanation of the X-ray background spectrum.

The process of e(+/-) pair production by relativistic nuclei on ambient photons is considered. The process is important for cosmic-ray nuclei in interstellar and intergalactic space as well as in galactic and extragalactic compact objects. The rate of this process is given by an integral of the cross section over the photon angular and energy distribution. In the case of isotropic photons, the angular integration is performed to provide an expression for the rate at given photon energy in the nucleus rest frame. The total rate then becomes a single integral of that rate over the photon energy distribution. Formulas are also given for the fractional energy loss of a relativistic nucleus colliding with a photon of a given energy in the rest frame. The nucleus energy-loss rate is integrated over the photon angular distribution in the case of isotropic photons, and simple fits are provided.

The recent observation by the detectors on board Granat of a spectral steepening above about 50 keV imposes a constraint on possible emission models for NGC 4151. This steepening, for example, is not well fitted by an exponential rollover characteristic of a purely thermal model, or by a Compton-downscattered power law. We find that this spectral behavior is consistent with that produced by a photon-starved nonthermal pair plasma with high compactness. This is the first quantitative test of the nonthermal pair plasma model against broad-band X-ray/gamma-ray observations. Above 200 keV or so, the nonthermal pair plasma model predicts an upturn in the spectrum, and a thermal annihilation feature around 511 keV. Such spectral behavior should be looked for with GRO.

The nonthermal e(+/-) pair model of the central engine of active galactic nuclei (AGNs) is discussed. The model assumes that nonthermal e(+/-) pairs are accelerated to highly relativistic energies in a compact region close to the central black hole and in the vicinity of some cold matter. The model has a small number of free parameters and explains a large body of AGN observations from EUV to soft gamma-rays. In particular, the model explains the existence of the UV bump, the soft X-rays excess, the canonical hard X-ray power law, the spectral hardening above about 10 keV, and some of the variability patterns in the soft and hard X-rays. In addition, the model explains the spectral steepening above about 50 keV seen in NGC 4151.

In the present study of the formation of steep soft X-ray excesses that are superposed on flatter, hard X-ray power-law spectra in nonthermal electron-positron pair cascade sources, the soft excess in pair-cascade AGN models appears as a steep power law superposed on the tail of the UV bump and the flat nonthermal (hard X-ray) power law. The model-parameter space in which an excess in soft X-rays is visible is ascertained, and the time-variability of soft excesses in pair cascade models is examined. It is established that the parameter space in which soft excesses appear encompasses the range of preferred input parameters for a recently development Compton reflection model of UV and X-ray emission from the central engine of an AGN.

Compton scattering of the microwave background photons by beams of ultrarelativistic electrons at large redshifts, z much greater than 1, is proposed as a source of some gamma-ray bursts. Such beams may be produced by cusps on superconducting cosmic strings. In the present model, a very narrow beam of ultrarelativistic electron-positron pairs scatters the microwave background photons into a very narrow beam of gamma rays. While the pairs lose energy, the beam of gamma rays opens up and sweeps the observer. The expected time variability and spectra of the resulting gamma-ray burst are calculated. The bursts have a rapid rise time, followed by a more gradual decline. The spectra are initially very hard and become softer during the decline.

We reconsider the recent model for the origin in the cosmic X-ray and gamma-ray background by Rogers and Field. The background in the model is due to an unresolved population of AGNs. An individual AGN spectrum contains three components: a power law with the energy index of alpha = 1.1, an enhanced reflection component, and a component from Compton scattering by relativistic electrons with a low energy cutoff at some minimum Lorentz factor, gamma(sub min) much greater than 1. The MeV bump seen in the gamma-ray background is then explained by inverse Compton emission by the electrons. We show that the model does not reproduce the shape of the observed X-ray and gamma-ray background below 10 MeV and that it overproduces the background at larger energies. Furthermore, we find the assumptions made for the Compton component to be physically inconsistent. Relaxing the inconsistent assumptions leads to model spectra even more different from that of the observed cosmic background. Thus, we can reject the hypothesis that the high-energy cosmic background is due to the described model.

It is shown here that reprocessing of radiation fron nonthermal pair cascades by cold material in the central parts of active galactic nuclei (AGN) gives rise to X-ray and gamma-ray spectra that satisfy current observational constraints. An average 1-30 keV X-ray spectral index alpha(x) of about 0.7 in the compact range 30-300 is obtained for a wide range of Lorentz factors of the injected electrons. The gamma-ray spectra are steep, with alpha(gamma) about two, and satisfy the observational constraints. Radiation from pair cascades exhibits steep power law decreases in soft X-rays similar to those observed in AGN. The overall picture is consistent with AGN having an accretion disk which intercepts and reprocesses a substantial fraction of the nonthermal continuum incident upon it from above and below.

Compton spectra produced by relativistic, unmagnetized, nonthermal electron-positron pairs injected into 'photon-starved' plasmas, where the luminosity in soft photons is much less than the power in the nonthermal pair injection, are studied. The overall form of the spectra consists of a rising spectrum at low energies, a single or double peak in the 0.1-1 MeV energy range, a cutoff above the peak, and a gamma-ray tail with a varying slope depending on the plasma parameters. The low-energy power law is mostly due to the primary soft photons being repeatedly upscattered by the hot thermal electrons. Comparisons are made with observations of compact objects.

Consideration is given to a model of gamma-ray burst sources based on repeated Compton scatterings of soft photons by relativistic nonthermal electrons. Relativistic electrons which are continuously produced in the source radiate the total power L supplied to them. Higher order Compton scatterings, which occur when L(soft) is much less than L make the model distinct. The spectrum having an X-ray energy index of about one results from the superposition of the spectral components from several orders of Compton scattering; the change of the slope at several hundred keV arises from the Klein-Nishina decline of the scattering cross section.

The model of gamma-ray-burst spectra proposed by Zdziarski and Lamb (1986, 1987) is reviewed and compared with observed spectra. In the model, the spectrum arises from multiple Compton scatterings in a nonthermal source with a deficit of soft photons. The steady-state electron distribution in such a source is a nonthermal power law that joins at low energies onto a thermal distribution that may be either optically thick or thin. Both cases lead naturally to a two-component spectrum with a photon-energy index in the X-ray spectrum which is much less than one and a photon-energy index in the gamma-ray spectrum which is larger than the X-ray photon-energy index. Fits to the observed gamma-ray-burst spectra (where data exist for both X-ray and gamma-ray energies) show good agreement with the model.