We report on Compton Gamma Ray Observatory observations of six detected pulsars: the Crab, Vela, Geminga, PSR B1509-58, PSR B1706-44, and PSR B1055-52. We combine these data with radio data and X-ray data to provide an overview of what is known about gamma-ray pulsars. We discuss light curves, spectra, and radio/gamma-ray phase offsets, and present several tentative patterns in the data. These include constant phase with γ-ray energy; a correlation between gamma-ray and X-ray luminosity; an anticorrelation between the gamma-ray luminosity and the efficiency in converting rotational energy loss into gamma-ray flux; and a correlation between the pulsar period and radio/gamma-ray phase offset. We also suggest that the emission models that have been proposed to date cannot explain the similarities of the average gamma-ray light curves observed over a wide range of energies. Further, unless a narrow beam is assumed, pulsars such as PSR B1055-52 and Geminga appear to be radiating a significant fraction of their rotational energy loss in the form of gamma rays.

Subject headings: gamma rays: observations — pulsars: general — radio continuum: stars — X-rays: stars

We discuss recent research on the structure and particle acceleration properties of relativistic shock waves in which the magnetic field is transverse to the flow direction in the upstream medium, and whose composition is either pure electrons and positrons or primarily electrons and positrons with an admixture of heavy ions. Particle-in-cell simulation techniques as well as analytic theory have been used to show that such shocks in pure pair plasmas are fully thermalized—the downstream particle spectra are relativistic Maxwellians at the temperature expected from the jump conditions. On the other hand, shocks containing heavy ions which are a minority constituent by number but which carry most of the energy density in the upstream medium do put ~20% of the flow energy into a nonthermal population of pairs downstream, whose distribution in energy space is N(E) ∝ E−2, where N(E)dE is the number of particles with energy between E and E + dE.

The mechanism of thermalization and particle acceleration is found to be synchrotron maser activity in the shock front, stimulated by the quasi-coherent gyration of the whole particle population as the plasma flowing into the shock reflects from the magnetic field in the shock front. The synchrotron maser modes radiated by the heavy ions are absorbed by the pairs at their (relativistic) cyclotron frequencies, allowing the maximum energy achievable by the pairs to be γ±m±c2 = mic2γ1/Zi, where γ1 is the Lorentz factor of the upstream flow and Zi, is the atomic number of the ions. The shock’s spatial structure is shown to contain a series of “overshoots” in the magnetic field, regions where the gyrating heavy ions compress the magnetic field to levels in excess of the eventual downstream value.

This shock model is applied to an interpretation of the structure of the inner regions of the Crab Nebula, in particular to the “wisps,” surface brightness enhancements near the pulsar. We argue that these surface brightness enhancements are the regions of magnetic overshoot, which appear brighter because the small Larmor radius pairs are compressed and radiate more efficiently in the regions of more intense magnetic field. This interpretation suggests that the structure of the shock terminating the pulsar’s wind in the Crab Nebula is spatially resolved, and allows one to measure γ1, and a number of other properties of the pulsar’s wind. We also discuss applications of the shock theory to the termination shocks of the winds from rotation-powered pulsars embedded in compact binaries. We show that this model adequately accounts for (and indeed predicted) the recently discovered X-ray flux from PSR 1957+20, and we discuss several other applications to other examples of these systems.

Subject headings: acceleration of particles — ISM: individual (Crab Nebula) — relativity — shock waves

Diffusive acceleration of electrons by the blast wave of SNR 1987A is shown to provide a reasonable fit to observations of the radio emission since it reappeared in 1990 July. As well as determining the diffusion coefficient in this model, the observations indicate that the compression ratio of the gas subshock embedded in the blast wave is ~2.7, much lower than the value of 4 expected of a strong shock front. We propose that protons accelerated by the shock front itself cause the weakening of the subshock by both heating and accelerating the upstream plasma, producing a structure with a precursor to the gas subshock. Preliminary calculations using a fluid model for the energetic particle population indicate that these effects can significantly weaken the subshock just a few years after the explosion, if the same rate of particle injection is assumed as in other models of the acceleration of cosmic rays.

Subject headings: acceleration of particles — ISM: individual (SNR 1987 A) — radio continuum: ISM — shock waves

Pulsed γ radiation from the Geminga pulsar was recorded by the COS B satellite from 50 MeV to 5 GeV between 1975 and 1982. It has been analysed to derive the source spectral properties as a function of time and phase. The two main peaks are separated by 0.50 ± 0.01 in phase. Significant pulsed emission has also been detected in both phase regions between the two main peaks. Significant spectral differences with phase have been found; the emission from the first peak is slightly softer than from the second one and the Interpeak 2 radiation is extremely soft. The first peak emission remained stable within the statistics over 7 years while the flux from the other phase intervals changed significantly with time. The phase dependence of the variability implies that all the observed source emission should be pulsed and that it consists of four discrete beams with different apertures and spectra. The beams characteristics are strikingly similar to those of four γ-ray beams generated by the Vela pulsar.

Subject headings: gamma rays: observations — pulsars: individual (Geminga, Vela)

IRAS observations the “filled-center” supernova remnant 3C 58 are used to derived upper limits of 0.4, 0.4, 0.8, and 1.5 Jy for its infrared flux density at 12, 25, 60, and 100 μm, respectively. These values imply a break in the spectrum of 3C58 above a few tens of gigahertz—two orders of magnitude below the break in the spectrum of the Crab Nebula—with a change of spectral index across the break for 3C 58 of more than 0.5. This is similar to previous results for another “filled-center” remnant, G74.9+1.2, from radio observations alone. This implies that their emission is not dominated by a constant injection rate of particles with a power-law distribution. Either the break is due to synchrotron losses in the past, with little particle injection in recent times or it is intrinsic to the acceleration mechanism at work (presumably from a central neutron star).

Subject headings: acceleration of particles — ISM: individual (3C 58) — radiation mechanisms: nonthermal — supernova remnants

The Compton Gamma-Ray Observatory (C-GRO) has completed a full-sky survey during which the number of known γ-ray pulsars has more than doubled. COMPTEL has observed the classical pulsars Crab and Vela on several occasions and has derived detailed pulse patterns and spectral parameters in the 0.7-30 MeV energy interval. The new C-GROγ-ray pulsars have different properties in terms of energy spectra and light-curve shapes, and, in fact, only the Crab is seen by all four C-GRO instruments. This raises intriguing questions about the particle acceleration processes and beaming taking place in the neutron magnetosphere. We have examined the COMPTEL data to add information on these objects in the 0.7-30 MeV energy interval and present evidence for the detection of one of them, PSR B1509-58. We have also undertaken a search for candidate radio pulsars whose ephemerides are well defined. The results of these analyses are presented.

Subject headings: gamma rays: observations — pulsars: general

Generally, people believe that the unpulsed high-energy gamma rays from the direction of Crab Nebula and pulsar comes from the nebula. But it is entirely possible that the unpulsed high-energy gamma rays from the direction of the Crab Nebula and pulsar are actually emitted from a region extending to a couple of light cylinder radii from the pulsar instead of from the extensive nebula. In this conference paper, we study the possibility that the unpulsed high-energy gamma rays from 100 MeV to 10 GeV are emitted from the extensive nebula. In our model, two pulsed photon beams from two different outer gaps cross each other beyond the light cylinder and result in pair production. Since the pitch angles of these pairs do not correlate with the local magnetic field, and the typical mean free path for pair production is comparable to the local radius of curvature, the subsequent synchrotron radiation and inverse-Compton scattering produce unpulsed X-rays and gamma-rays respectively.

Subject headings: acceleration of particles — gamma rays: theory — ISM: individual (Crab Nebula) — pulsars: individual (Crab) — radiation mechanisms: nonthermal

We have investigated the electrodynamics of magnetic neutron stars accreting from Keplerian disks and the implications for particle acceleration and γ-ray emission by such systems. We argue that the particle density in the magnetospheres of such stars is larger by orders of magnitude than the Goldreich-Julian density, so that the formation of vacuum gaps is unlikely. We show that even if the star rotates slowly, electromotive forces ( EMFs ) of order 1015 V are produced by the interaction of plasma in the accretion disk with the magnetic field of the neutron star. The resistance of the disk-magnetosphere-star circuit is small, and hence these EMFs drive very large conduction currents. Such large currents are likely to produce magnetospheric instabilities, such as relativistic double layers and reconnection events, that can accelerate electrons or ions to very high energies.

Subject headings: acceleration of particles — accretion, accretion disks — MHD — stars: neutron

Interaction of an accretion disk with the magnetic field of a neutron star produces large electromotive forces, which drive large conduction currents in the disk-magnetosphere-star circuit. Here we argue that such large conduction currents will cause microscopic and macroscopic instabilities in the magnetosphere. If the minimum plasma density in the magnetosphere is relatively low (≲109 cm−3 ), current-driven micro-instabilities may cause relativistic double layers to form, producing voltage differences in excess of 1012 V and accelerating charged particles to very high energies. If instead the plasma density is higher (≳ 109 cm−3 ), twisting of the stellar magnetic field is likely to cause magnetic field reconnection. This reconnection will be relativistic, accelerating plasma in the magnetosphere to relativistic speeds and a small fraction of particles to very high energies. Interaction of these high-energy particles with X-rays, γ-rays, and accreting plasma may produce detectable high-energy radiation.

Subject headings: acceleration of particles — accretion, accretion disks — gamma rays: theory — plasmas — radiation mechanisms: nonthermal — stars: neutron

Recent numerical models for SNR evolution are presented, including first-order Fermi acceleration with injection of suprathermal particles at the shock wave, heating due to dissipation of Alfvén waves in the precursor region and radiative cooling of the thermal plasma. The X-ray fluxes obtained from these SNR models show significant differences depending on the acceleration efficiency of cosmic rays. γ-ray fluxes are calculated originating from π0-decay of pions generated by collisions of the high-energy particles with the thermal plasma. Cooling of the thermal plasma and dissipation of Alfvén waves in the precursor are important to determine the final amount of the explosion energy ESN which is transferred into cosmic rays.

Subject headings: acceleration of particles — cosmic rays — gamma rays: theory — shock waves — supernova remnants

I present model calculations of profiles and two-dimensional images of the radio synchrotron emission of young supernova remnants, concentrating on observable effects of relativistic electrons diffusing upstream of the shock wave. If the preshock electron scattering mean free path is sufficiently long, observable synchrotron halos outside the bulk of the radio emission can potentially result; their absence can constrain the mean free path from above. If scattering is primarily due, as expected, to Alfvén waves with amplitude δB, the halo is expected to extend a distance of order rg c(δB/B)−2 /vs beyond the shock, where rg is the gyroradius of the electrons emitting at the observed frequency, B is the upstream magnetic field strength, vs is the shock velocity, and the amplitude δB refers to waves with wavelength comparable to rg , of order 1013 cm for typical supernova-remnant parameters. However, the detailed geometry of the halo varies with the assumptions about particle acceleration in the shock wave. I present an atlas of model profiles and images as a function of preshock diffusion length, of aspect angle between the magnetic field and the line of sight, and of other relevant parameters.

Subject headings: radiation mechanisms: miscellaneous — shock waves — supernova remnants

The recent observational data on gamma-ray bursts are reviewed. Burst time histories display features at the millisecond level which suggest a compact object origin. Lines in the energy spectra and spectral evolution point toward a Galactic neutron star origin, even though line features have not yet been confirmed in recent data. The Galactic distribution of burst sources, however, is both isotropic and sampled to its characteristic distance, making it unlikely that the sources are related to populations of Galactic neutrons stars we are familiar with. Counterpart searches, previously carried out years after the gamma-ray bursts, are now proceeding days after the events. Based on the current data, it is impossible to conclude whether bursters are Galactic, extragalactic, or both. Data being returned from current experiments, as well as data from new experiments to be launched in the next few years, may yet provide the answer.

Subject heading: gamma rays: bursts

Gamma-ray bursts have always been intriguing sources to study in terms of particle acceleration, but not since their discovery two decades ago has the theory of these objects been in such turmoil. Prior to the launch of Compton Gamma-Ray Observatory and observations by BATSE, there was strong evidence pointing to magnetized Galactic neutron stars as the sources of gamma-ray bursts. However, since BATSE the observational picture has changed dramatically, requiring much more distant and possibly cosmological sources. I review the history of gamma-ray burst theory from the era of growing consensus for nearby neutron stars to the recent explosion of halo and cosmological models and the impact of the present confusion on the particle acceleration problem.

Subject headings: acceleration of particles — gamma rays: bursts

Is it shown that for those astronomical models of cosmic gamma-ray bursts (GRBs) which are associated with galactic neutron stars (NSs), the initial energy of the outburst could be converted to gamma-rays through processes of particle acceleration. The main emission mechanisms are considered for two basic alternatives, when particles are accelerated either by radiative pressure or by an electric field.

Subject headings: acceleration of particles — gamma rays: bursts — radiation mechanisms: nonthermal

Many proficient gamma-ray sources show energy spectra that are consistent with E−2 primary spectra. Such sources include recently identified gamma-ray quasars and some gamma-ray bursts. Assuming thick target conversion, this is consistent with shock acceleration, and the dominance of the gamma rays of the luminosity is also consistent with previous predictions of high production efficiency of fresh cosmic rays in shocks. The spectral cutoffs in the gamma rays may offer clues as to whether the high-energy particles are electrons or protons. Resolution of this matter might have implications for the nature of the sources and for theory of shock accelerated electrons.

Subject headings: acceleration of particles — gamma rays: bursts — shock waves

In 1983 Samorski and Stamm found evidence of ultra-high-energy (UHE) gamma-ray emission from Cygnus X-3 in archival data from the Kiel air shower array. The emission appeared to be emitted only at one phase in the binary orbit. The Kiel array was sensitive to air showers generated by particles with energies above 3 × 1015 eV. Their discovery led to searches in archival data from other arrays for evidence of UHE emission from Cygnus X-3 and other potential sources in the Galaxy, and to the construction of new air shower arrays dedicated to UHE astronomy.

In the mid 1980s many groups around the world claimed detections of Cygnus X-3, Hercules X-1, Vela X-1, and other objects. Cygnus X-3 is strongly variable at all wavelengths, and so it would be surprising if this variability did not extend to UHE gamma-rays also. This indeed appeared to be the case, and in the mid to late 1980s the UHE gamma-ray output showed a steady decline apart from transient UHE emission following giant radio flares. With the exception of possible detections above 1017 eV by the Fly’s Eye and Akeno Array, I am unaware of any statistically compelling claimed detections of this source reported over the past 4 years. The new arrays—CASA-MIA and SPASE—have not detected any of the claimed sources, and it seems likely that at least some of these early claims may have been premature. However, although none of the observations taken alone was of extremely high statistical significance, the observations of Cygnus X-3 by different groups showed a consistency which I feel cannot be ascribed to chance alone. The observations of Cygnus X-3 and the Crab Nebula/Pulsar are reviewed with particular reference to the results published during the last 5 years.

Subject headings: gamma rays: observations — pulsars: individual (Crab Nebula) — stars: individual (Cygnus X-3)

A short-lived (weeks-months) burst of radio emission, assumed to be synchrotron radiation, is sometimes associated with objects detected as soft X-ray transients. The “synchrotron bubble model” which has been applied to these radio transients assumes that the source is initially optically thick, and the model therefore takes no account of the details of the electron acceleration required. The initial increase in the radio flux is attributed to decreasing absorption in the expanding source. Observations of the soft X-ray transient GRS 1124 – 68 (X-ray Nova Muscae 1991) showed, for the first time, the rising phase of such a radio transient. Another radio source, the Galactic Center Transient, was discovered serendipitously in the radio and observed throughout its rise and decay. In both cases the radio flux was a decreasing function of frequency throughout the periods of increasing and decreasing flux, which is not consistent with the optically thick assumption of the synchrotron bubble model. When this assumption is discarded the mechanism responsible for producing the ultrarelativistic electrons, or at the very least the rate at which those electrons become available, must be considered. A generalization of the simple model to include a source which supplies relativistic electrons with a constant energy spectrum is discussed. While such a model can go some way toward matching the observations, it is suggested that the time is ripe for an explicit model of the electron acceleration mechanism. Some of the features such a model may have are presented.

Subject headings: acceleration of particles — radio continuum: general — radiation mechanisms: nonthermal — X-rays: bursts

The γ-ray burst brightness distribution is inhomogeneous and the distribution on the sky is nearly isotropic. These features argue against an association of γ-ray bursts with those Galactic objects that are known to exhibit a strong concentration toward the Galactic center or plane. The observed statistical properties indicate a cosmological origin. Circumstantial evidence suggests that neutron stars are involved in the burst phenomenon. Here we consider Population II neutron stars in an extended Galactic Halo (EGH) as an alternative to cosmological scenarios. The BATSE data indicate a small deviation from isotropy near the 2 σ level of statistical significance. If confirmed for an increasing number of bursts, these anisotropies could rule out cosmological scenarios. On the other hand, EGH models require small anisotropies like those observed by BATSE. We consider simple distribution models to determine the generic properties such halos must have to be consistent with the observations and discuss the implications of the corresponding distance scale on burst models.

Subject headings: gamma rays: bursts — stars: neutron — stars: Population II — stars: statistics

The principal discovery of the Compton Gamma-Ray Observatory (GRO) concerning gamma-ray bursts is that these sources are isotropic but with a comparative deficiency of fainter sources, suggesting that they are probably cosmological in origin. If they are at such large distances from Earth then they are extremely luminous and compact. A consequence of this is that two-photon pair production attenuation of the gamma-ray continuum cannot be avoided unless the source radiation is substantially beamed. Most sources do not display gamma-ray turnovers although a few GRB detected by GRO exhibit distinct spectral breaks in the MeV range. A derivation of the relationship defining of the degree of beaming in burst sources with spectral breaks due to γ-γ attenuation, as a function of source spectral index and break energy, is presented. It is found that sources at distances of ~1 Gpc must typically be beamed with bulk Lorentz factors of around 103 -104 , indicating powerful bulk acceleration in bursts, although these Lorentz factors are reduced markedly for steep source spectra. Since the source spectra are not strongly Comptonized, such beaming will blueshift the γ-γ attenuation breaks to energies much higher than 1 MeV; an absolute lower bound to the source bulk Lorentz factor is determined from this additional constraint. This blueshifting suggests that those sources with MeV breaks may not be cosmological, or that their breaks are produced by a mechanism that dominates γ-γ attenuation at these energies.

Subject headings: gamma-rays: bursts — radiation mechanisms: nonthermal

Since several gamma-ray bursts seem to have their origin in the vicinity of neutron stars, where strong magnetic fields are present (B ≃ 108T), electrons and positrons have to be described, in this instance, by relativistic Landau states. As is well known, strong magnetic fields produce important effects: motion perpendicular to the field is quantized, transverse momentum is not conserved, and polarization effects become important. Moreover, in contrast to the field-free case, exotic processes such as 1γ pair-annihilation are not forbidden. With growing magnetic field, the cross section of this process becomes more and more nonnegligible. This cross section strongly depends on the polarization of the electron-positron pairs. Thus the deceleration via annihilation turns out to be a strongly polarization-dependent phenomenon in magnetized electron-positron plasmas.

Subject headings: acceleration of particles — gamma rays: theory — polarization — radiation mechanisms: nonthermal — stars: neutron