Search

The unprecedented spatial resolution of the Chandra observatory opens the possibility to detect with relatively high accuracy proper motions at X-ray wavelengths. We have conducted an astrometric study of three of the "Magnificent Seven", the thermally emitting radio quiet isolated neutron stars (INSs) discovered by ROSAT. These three INSs (RX J0420.0-5022, RX J0806.4-4123 and RX J1308.6+2127/RBS 1223) either lack an optical counterpart or have one too faint to be used for astrometric purposes. We obtained ACIS observations 3 to 5 years apart to constrain or measure the displacement of the sources on the X-ray sky using as reference the background of extragalactic or remote galactic X-ray sources. Upper limits of 138 mas/yr and 76 mas/yr on the proper motion of RX J0420.0-5022 and RX J0806.4-4123, respectively, have already been presented in Motch et al. (2007). Here we report the very significant measurement (~ 10 sigma) of the proper motion of the third INS of our program, RX J1308.6+2127/RBS1223. Comparing observations obtained in 2002 and 2007 reveals a displacement of 1.1 arcsec implying a yearly proper motion of 223 mas, the second fastest measured for the ROSAT discovered INSs. The source is rapidly moving away from the galactic plane at a speed which precludes any significant accretion of matter from the interstellar medium. Its transverse velocity of ~ 740 (d/700pc) km/s might be the largest of the "Magnificent Seven" and among the fastest recorded for neutron stars. RX J1308.6+2127/RBS1223 is thus a young high velocity cooling neutron star. The source may have its origin in the closest part of the Scutum OB2 association about 0.8 Myr ago, an age consistent with that expected from cooling curves, but significantly younger than inferred from pulse timing measurements (1.5 Myr)., Comment: 3 pages, 2 figures, proceedings of the conference "40 Years of Pulsars", 12-17 August 2007, Montreal, Canada

We present a brief review of Chandra X-ray Observatory observations of neutron stars. The outstanding spatial and spectral resolution of this great observatory have allowed for observations of unprecedented clarity and accuracy. Many of these observations have provided new insights into neutron star physics. We present an admittedly biased and overly brief overview of these observations, highlighting some new discoveries made possible by the Observatory's unique capabilities. We also include our analysis of recent multiwavelength observations of the putative pulsar and its pulsar-wind nebula in the IC 443 SNR., Comment: 10 pages, 14 figures. Submitted to Astronomy and Space Science. Based on presentation at Isolated Neutron Stars, London, 2006

The superb spatial resolution of Chandra has allowed us to detect a 20''-long tail behind the Geminga pulsar, with a hard spectrum (photon index 1.0+/-0.2) and a luminosity (1.3+/-0.2) 10^{29} ergs/s in the 0.5 - 8 keV band, for an assumed distance of 200 pc. The tail could be either a pulsar jet, confined by a toroidal magnetic field of about 100 microGauss, or it can be associated with the shocked relativistic wind behind the supersonically moving pulsar confined by the ram pressure of the oncoming interstellar medium. We also detected an arc-like structure 5'' - 7'' ahead of the pulsar, extended perpendicular to the tail, with a factor of 3 lower luminosity. We see a 3-sigma enhancement in the Chandra image apparently connecting the arc with the southern outer tail that has been possibly detected with XMM-Newton. The observed structures imply that the Geminga's pulsar wind is intrinsically anisotropic., Comment: Revised version: data analysis described in more detail, Figure 2 replaced; 6 pages, 2 color figures; accepted for publication in ApJ (v.643, 2006 June 1)

We observed the gamma-ray pulsar Geminga with the HST STIS/MAMA detectors to measure the Geminga's UV spectrum and pulsations. The slope of the far-UV (FUV) spectrum is close to that of a Rayleigh-Jeans spectrum, suggesting that the FUV radiation is dominated by thermal emission from the neutron star (NS) surface. The measured FUV flux, F_FUV=(3.7+/-0.2)x10^-15 ergs cm^-2 s^-1 in 1155-1702 A band, corresponds to a brightness temperature T_RJ=(0.3-0.4)(d_200/R_13)^2 MK. The soft thermal component of the Geminga's X-ray spectrum measured with the XMM-Newton observatory corresponds to a temperature T_s=0.49+/-0.01 MK and radius R_s = (12.9+/-1.0)d_200 km. Unlike other NSs detected in the UV-optical, the FUV spectrum of Geminga lies below the extrapolation of the soft thermal component, which might be associated with the Geminga's very low temperature. Surprisingly, the thermal FUV radiation is strongly pulsed, showing a narrow dip at a phase close to that of a broader minimum of the soft X-ray light curve. The strong pulsations might be attributed to partial occultations of the thermal UV radiation by regions of the magnetosphere filled with electron/positron plasma. In contrast with the FUV spectrum, the near-IR (NIR) through near-UV (NUV) spectrum of Geminga is clearly nonthermal. It can be described by a power-law model, F_nu \propto nu^{-alpha} with a slope alpha = 0.43+/-0.15, close to alpha = 0.56+/-0.24 of the hard X-ray (E>2.5 keV) magnetospheric component. The observed NIR-optical-NUV spectrum is marginally consistent with the extrapolation of the X-ray magnetospheric spectrum. The NUV pulsations, however, do not show a clear correlation with the hard X-ray pulsations., Comment: 17 pages including 14 figures, ApJ, 625 in press

We obtained deep optical imaging of the thermally emitting X-ray bright and radio-quiet isolated neutron star RX J1605.3+3249 with the Subaru telescope in 1999 and 2003. Together with archival HST images acquired in 2001 these data reveal a proper motion of mu = 144.5 +/- 13.2 mas/yr. This implies a relatively high spatial velocity and indicates that the star is unlikely to be re-heated by accretion of matter from the interstellar medium. Assuming that RX J1605.3+3249 is a young (10^5-10^6 yr) cooling neutron star, its apparent trajectory is consistent with a birth in the nearby Sco OB2 OB association at a location close to that derived for RX J1856.5-3754 and perhaps also to that of RX J0720.4-3125. This suggests that the X-ray bright part of ROSAT-discovered isolated neutron stars is dominated by the production of the Sco OB2 complex which is the closest OB association and a part of the Gould belt. The B and R magnitudes of the faint optical counterpart did not vary from 1999 to 2003 at B = 27.22 +/- 0.10. Its B-R colour index of +0.32 +/- 0.17 is significantly redder than that of other isolated neutron stars and the optical flux lies a factor 11.5 above the extrapolation of the X-ray blackbody-like spectrum. The red optical colour reveals the presence of an additional emitting component in the optical regime over the main neutron star thermal emission. We also discovered a small elongated Halpha nebula approximately centered on the neutron star and aligned with the direction of motion. The width of the nebula is unresolved and smaller than ~ 0.4" for a length of about 1". The shape of the Balmer emitting nebula is very different from those seen close to other neutron stars and should be confirmed by follow-up observations. We shortly discuss the possible mechanisms which could give rise to such a geometry., Comment: 9 pages, 6 figures, resolution of some figures reduced to meet astro-ph file size restriction, accepted for publication in A&A

We report on the analysis of new X-ray data obtained with XMM-Newton and Chandra from two ROSAT-discovered X-ray dim isolated neutron stars. RX J0806.4-4123 was observed with XMM-Newton in April 2003, 2.5 years after the first observation. The EPIC-pn data confirm that this object is an X-ray pulsar with 11.371 s neutron star spin period. The X-ray spectrum is consistent with absorbed black-body emission with a temperature kT = 96 eV and NH = 4x10^19 cm^-2 without significant changes between the two observations. Four XMM-Newton observations of RX J0420.0-5022 between December 2002 and July 2003 did not confirm the 22.7 s pulsations originally indicated in ROSAT data, but clearly reveal a 3.453 s period. A fit to the X-ray spectrum using an absorbed black-body model yields kT = 45 eV, the lowest value found from the small group of XDINs and NH = 1.0x10^20 cm^-2. Including a broad absorption line improves the quality of the spectral fits considerably for both objects and may indicate the presence of absorption features similar to those reported from RBS1223, RX J1605.3+3249 and RX J0720.4-3125. For both targets we derive accurate X-ray positions from the Chandra data and present an optical counterpart candidate for RX J0420.0-5022 with B = 26.6+/-0.3 mag from VLT imaging., Comment: 11 pages, 7 figures, accepted for publication in A&A

We present the timing and spectral analyses of theXMM-newton data on the 17-Myr-old, nearby radio pulsar B0950+08. This observation revealed pulsations of the X-ray flux of the pulsar at its radio period. The pulse shape and pulsed fraction are apparently different at lower and higher energies of the observed 0.2-10 keV energy range, which suggests that the radiation cannot be explained by a single emission mechanism. The X-ray spectrum of the pulsar can be fitted with a power-law model with a photon index about 1.75 and an (isotropic) luminosity about 9.8e29 erg/s in the 0.2-10 keV. Better fits are obtained with two-component, power-law plus thermal, models with index of 1.30 and 9.7e29 erg/s for the power-law component that presumably originates from the pulsar's magnetosphere. The thermal component, dominating at E>0.7 keV, can be interpreted as radiation from heated polar caps on the neutron star surface covered with a hydrogen atmosphere. The inferred effective temperature, radius, and bolometric luminosity of the polar caps are about 1 MK, 250 m, and 3e29 erg/s. Optical through X-ray nonthermal spectrum of the pulsar can be described as a single power-law with index 1.3-1.4 for the two-component X-ray fit. The ratio of the nonthermal X-ray (1-10 keV) luminosity to the nonthermal optical (4000-9000 \AA) luminosity is within the range of 1e2-1e3 observed for younger pulsars, which suggests that the magnetospheric X-ray and optical emissions are powered by the same mechanism in all pulsars. An upper limit on the temperature of the bulk of the neutron star surface, inferred from the optical and X-ray data, is about 0.15 MK. We also analyze X-ray observations of several other old pulsars, B2224+65, J2043+2740, B0628-28, B1813-36, B1929+10, and B0823+26., Comment: To be published in ApJ. Nonthermal optical and X-ray luminosities of seven radio pulsars are updated and presented in a new Table. Figure 6 showing the ratios of the luminosities vs. spin-down energy is also updated

We discuss recent Chandra and XMM-Newton observations of the bright isolated neutron star RX J1856.5-3754 and suggest that the absence of any line features is due to effects of a high magnetic field strength (~10^13 G). Using different models for the temperature distribution across the neutron star surface assuming blackbody emission to fit the optical and X-ray spectrum and we derive a conservative lower limit of the "apparent" neutron star radius of 16.5 km x (d/117 pc). This corresponds to the radius for the "true" (de-redshifted) radius of 14 km for a 1.4 Msun neutron star, indicating a stiff equation of state at high densities. A comparison of the result with mass-radius diagrams shows that quark stars and neutron stars with quark matter cores can be ruled out with high confidence., Comment: 6 page, 2 figures, "The Restless High-Energy Universe" Proceedings of the symposium dedicated to six years of successful BeppoSAX operations Amsterdam, May 5-8, 2003

The XMM-Newton spectra of the isolated neutron star RX J0720.4-3125 show deviations from a Planckian energy distribution below 400 eV, similar to the spectra of RBS1223, another long-period X-ray pulsar, as reported recently by Haberl et al. (2003). For a Gaussian-shaped absorption line we derive an energy of 271 eV and an equivalent width of -40 eV from the phase-averaged spectra of RX J0720.4-3125. We investigate the spectral variations seen in hardness ratios as function of pulse phase and find that they are best described by changes in the depth of the absorption line. The line equivalent width changes between -31 eV around intensity maximum of the pulse and -58 eV at the declining part of the pulse. Small variations (<20 eV) of the line energy with pulse phase may still be caused by statistical fluctuations. On the other hand, the black-body temperature varies significantly by 2.5 eV (statistical 90% errors typically 0.7 eV) reaching the highest value at pulse maximum. One possible interpretation for the absorption line is cyclotron resonance scattering of protons in a magnetic field with B about 5x10^13 G. This field strength is compatible with estimates inferred from recent spin down measurements of the pulsar., Comment: Accepted for publication in A&A

X-ray observations of middle-aged pulsars allow one to study nonthermal radiation from pulsar magnetospheres and thermal radiation from neutron star (NS) surfaces. In particular, from the analysis of thermal radiation one can infer the surface temperatures and radii of NSs, which is important for investigating evolution of these objects and constraining the equation of state of the superdense matter in the NS interiors. Here we present results of XMM observations of three middle-aged pulsars, J0538+2817, B0656+14 and J0633+1746 (Geminga), and briefly discuss mechanisms of their X-ray emission., Comment: 6 pages, 8 figures; to be published in Memorie della Societa' Astronomica Italiana, the Proceedings of the EPIC Consortium (held on Oct 14-16, 2003 in Palermo)

The X-ray source 1E 1207.4-5209 is a compact central object in the G296.5+10.0 supernova remnant. Its spin period of 424 ms, discovered with the Chandra X-ray Observatory, suggests that it is a neutron star. The X-ray spectrum of this radio-quiet pulsar shows at least two absorption lines, first spectral features discovered in radiation from an isolated neutron star. Here we report the results of timing analysis of Chandra and XMM-Newton observations of this source showing a non-monotonous behavior of its period. We discuss three hypotheses which may explain the observational result. The first one assumes that 1E 1207.$-5209 is a glitching pulsar, with frequency jumps of \Delta f > 5 \muHz occurring every 1-2 years. The second hypothesis explains the deviations from a steady spin-down as due to accretion, with accretion rate varying from \sim 10^{13} to >10^{16} g s^{-1}, from a disk possibly formed from ejecta produced in the supernova explosion. Finally, the period variations could be explained assuming that the pulsar is in a wide binary system with a long period, P_orb \sim 0.2-6 yr, and a low-mass companion, M_2 < 0.3 M_\odot., Comment: 20 pages, 5 figures, accepted for publications in ApJ. 2004 ApJ, in press

Observations of thermal radiation from neutron stars allow one to measure the surface temperatures and confront them with cooling scenarios. Detection of gravitationally redshifted spectral lines can yield the mass-to-radius ratio. In the few cases when the distance is known, one can measure the neutron star radius, which is particularly important to constrain the equation of state of the superdense matter in the neutron star interiors. Finally, one can infer the chemical composition of the neutron star surface layers, which provides information about formation of neutron stars and their interaction with the environments. We present the observational results on thermal radiation from active pulsars and radio-quiet neutron stars, with emphasis on the results obtained with the Chandra X-ray Observatory and discuss some implications of these results. In particular, we present a new interpretation of the thermal emission detected in the broad energy range (from optical to X-rays) from the puzzling object RX J1856.5-3754, Comment: To be published in the Proceedings of the XXI Texas Symposium on Relativistic Astrophysics

ESO 4m class telescope and VLT deep imaging of the isolated neutron star RX J0720.4-3125 reveals a proper motion of mu = 97 +/-12 mas/yr and a blue U-B color index. We show that a neutron star atmosphere model modified to account for a limited amount of hydrogen on the star's surface can well represent both the optical and X-ray data without invoking any additional components. The large proper motion almost completely excludes the possibility that accretion from interstellar medium is the powering mechanism of the X-ray emission. It also implies that the proposed spin down is entirely due to magnetic dipole losses. RX J0720.4-3125 is thus a very likely middle aged cooling neutron star. Its overall properties are quite similar to some of the long period radio pulsars recently discovered, giving further support to the idea that RX J0720.4-3125 may be a pulsar whose narrow radio beam does not cross the Earth., Comment: Submitted to Astronomy and Astrophysics, 8 pages 5 figures

We present results of the analysis of data collected in 57-ks XMM-Newton and 505-ks Chandra observations of the nearby ~= 120 pc isolated neutron star RX J1856.5-3754. We confirm most of the statements made by Burwitz et al. (2001) who discussed the original 55-ks Chandra data. Detailed spectral analysis of the combined X-ray and optical data rules out the currently available nonmagnetic light and heavy element neutron star atmosphere (LTE) models with hydrogen, helium, iron and solar compositions. We find that strongly magnetized atmosphere models also are unable to represent the data. The X-ray and optical data show no spectral features and are best fitted with a two-component blackbody model with kT_{bb,X}^{\infty} ~= 63.5 eV and R_{bb,X}^{\infty} ~= 4.4(d/120 pc) km for the hot X-ray emitting region, and kT_{bb,opt}^{\infty}<33 eV and R_{bb,opt}^{\infty}>17(d/120 pc) km for the rest of the neutron star surface responsible for the optical flux. The large number of counts collected with XMM-Newton allows us to reduce the upper limit on periodic variation in the X-ray range down to 1.3% (at a 2\sigma confidence level) in the 10^{-3}-50 Hz frequency range. In an attempt to explain this small variability, we discuss an one-component model with kT_{bb}^{\infty} ~= 63 eV and R_{bb}^{\infty} ~= 12.3(d/120 pc) km. This model requires a low radiative efficiency in the X-ray domain, which may be expected if the neutron star has a condensed matter surface., Comment: 6 pages, (including 3 figures) accepted for publication in Astronomy and Astrophysics (A&A)

We observed 1E 1207.4--5209, a neutron star in the center of the supernova remnant PKS 1209--51/52, with the ACIS detector aboard the Chandra X-ray observatory and detected two absorption features in the source spectrum. The features are centered near 0.7 keV and 1.4 keV, their equivalent widths are about 0.1 keV. We discuss various possible interpretations of the absorption features and exclude some of them. A likely interpretation is that the features are associated with atomic transitions of once-ionized helium in the neutron star atmosphere with a strong magnetic field. The first clear detection of absorption features in the spectrum of an isolated neutron star provides an opportunity to measure the mass-to-radius ratio and constrain the equation of state of the superdense matter., Comment: 11 pages, including 1 figure and 1 table, to be published in ApJL

Models of thermal emission of neutron stars, presumably formed in their atmospheres, are needed to infer the surface temperatures, magnetic fields, chemical composition, and neutron star masses and radii from the observational data. This information, supplemented with model equations of state and neutron star cooling models, is expected to move us further in understanding the fundamental properties of the superdense matter in the neutron star interiors. The neutron star atmospheres are very different from those of usual stars due to the immense gravity and huge magnetic fields. In this presentation we review the current status of the neutron star atmosphere modeling and present most important results., Comment: 10 pages, 13 color figures; to be published in Proceedings of the 270. Heraeus Seminar on Neutron Stars, Pulsars and Supernova Remnants, Bad Honnef (Germany), Jan. 21-25, 2002, eds. W. Becker, H. Lesch and J. Truemper

The outstanding capabilities of the Chandra X-ray observatory have greatly increased our potential to observe and analyze thermal radiation from the surfaces of neutron stars (NSs). Such observations allow one to measure the surface temperatures and confront them with the predictions of the NS cooling models. Detection of gravitationally redshifted spectral lines can yield the NS mass-to-radius ratio. In rare cases when the distance is known, one can measure the NS radius, which is particularly important to constrain the equation of state of the superdense matter in the NS interiors. Finally, one can infer the chemical composition of the NS surface layers, which provides information about formation of NSs and their interaction with the environment. We overview the recent Chandra results on the thermal radiation from various types of NSs -- active pulsars, young radio-quiet neutron stars in supernova remnants, old radio-silent ``dim'' neutron stars -- and discuss their implications., Comment: URL changed for Figures 1, 12 and 18: ftp://ftp.xray.mpe.mpg.de/people/zavlin

The isolated neutron star RX J0806.4-4123 was observed with XMM-Newton in November 2000. The data from the three EPIC instruments allowed us (i) to derive an improved X-ray position to an accuracy of 2-3'', (ii) to accumulate the first medium-resolution soft X-ray spectra of high statistical quality and (iii) to find a candidate for the neutron star rotation period. Although this period of 11.3714 s is formally detected at a 3.5-\sigma level in the EPIC-pn data, the similar pulse profiles deduced from all three EPIC instruments increase the confidence that the period is real. The pulsed fraction of ~6% would then be the weakest X-ray flux modulation detected from dim isolated neutron stars. We fitted the X-ray spectra with blackbody and neutron star atmosphere models and discuss the results with respect to the brightness limit placed by optical images. The reduced size of the error circle on the X-ray position should allow deeper searches for an optical counterpart., Comment: 7 pages, 6 figures, accepted by A&A

The pn camera of EPIC is inherently robust against radiation damage effects. Nevertheless, the knowledge of charge transfer inefficiency (CTI) of the pn camera is crucial for obtaining the correct energy scale. We describe detailed in-orbit monitoring of this effect, utilizing the internal calibration source. We find that during the first two years in orbit the CTI increased by about 4% for Al-K and by about 7% for Mn-K. The increases in the CTI are well within expectations, with no measurable effect on the energy resolution., Comment: 4 pages, 10 figures; to appear in the Proceedings of the Symposium on `New Visions of the X-ray Universe in the XMM-Newton and Chandra Era', 26-30 November 2001, ESTEC, The Netherlands

Second Chandra observation of 1E 1207.4-5209, the central source of the supernova remnant PKS 1209-51/52, allowed us to confirm the previously detected period of 424 ms and, assuming a uniform spin-down, estimate the period derivative, Pdot (0.7-3)e-14 s/s. The corresponding characteristic age of the pulsar, P/2Pdot~200-900 kyr, is much larger than the estimated age of the SNR, \~7 kyr. The values of the spin-down luminosity, Edot (0.4-1.6)e34 erg/s, and conventional magnetic field, B (2-4)e12 G, are typical for a middle-aged radio pulsar, although no manifestations of pulsar activity have been observed. If 1E 1207.4-5209 is indeed the neutron star formed in the same supernova explosion that created PKS 1209-51/52, such a discrepancy in ages could be explained either by a long initial period, close to its current value, or, less likely, by a very large braking index of the pulsar. Alternatively, the pulsar could be a foreground object unrelated to the supernova remnant, but the probability of such a coincidence is very low., Comment: 4 pages, 3 figures, to appear in ApJ Letters, v. 569 (April 2002)

One of the three cameras behind the X-ray telescopes on board XMM-Newton is equipped with a pn CCD detector. Here the current status of the pn spectral response calibration is summarized. Several parameters describing the spectral response of the EPIC-pn detector show spatial dependencies. E.g. spectral resolution and single- to double-pixel event ratios are functions of the charge transfer losses and therefore depend on the distance to the read-out node of the CCD (the RAWY detector coordinate)., Comment: 4 pages, 6 figures, To appear in the Proceedings of the Symposium on `New Visions of the X-ray Universe in the XMM-Newton and Chandra Era', 26-30 November 2001, ESTEC, The Netherlands

We report on spectral and timing observations of the nearest millisecond pulsar J0437-4715 with Chandra. The pulsar spectrum, detected up to 7 keV, cannot be described by a simple one-component model. We suggest that it consists of two components, a nonthermal power-law spectrum generated in the pulsar magnetosphere, with a photon index about 2, and a thermal spectrum emitted by heated polar caps, with a temperature decreasing outwards from 2 MK to 0.5 MK. The lack of spectral features in the thermal component suggests that the neutron star surface is covered by a hydrogen (or helium) atmosphere. The timing analysis shows one X-ray pulse per period, with a pulsed fraction of about 40% and the peak at the same pulse phase as the radio peak. No synchrotron pulsar-wind nebula is seen in the X-rays., Comment: to appear in ApJ; 13 pages, 5 figures (1 in color)

X-ray images of some young supernova remnants show bright point sources which have not been detected in radio, optical and gamma-ray bands. Despite the similarity of the X-ray spectra of these objects, they show a variety of temporal properties. Most likely, they are neutron stars whose properties (spin periods? magnetic fields? environments?) are different from those of radio and/or gamma-ray pulsars. We present an overview of observational results on several objects of this class -- the central sources of Cassiopeia A, RX J0852-4622, RCW 103, Puppis A, and PKS 1209-51/52 -- with emphasis on the recent Chandra observations., Comment: 10 pages, uses newpasp.sty. To appear in "Neutron Stars in Supernova Remnants" (ASP Conference Proceedings), eds P.O. Slane and B.M. Gaensler

We report the results of the spectral and timing analysis of observations of the Vela pulsar with the Chandra X-ray Observatory. The spectrum shows no statistically significant spectral lines in the observed 0.25--8.0 keV band. It consists of two distinct continuum components. The softer component can be modeled as either a magnetic hydrogen atmosphere spectrum with kT = 59 +- 3 eV, R = 15.5 +- 1.5 km, or a standard blackbody with kT = 129 +- 4 eV, R = 2.1 +- 0.2 km (the radii are for a distance of 250 pc). The harder component, modeled as a power-law spectrum, gives photon indices depending on the model adopted for the soft component: gamma = 1.5 +- 0.3 for the magnetic atmosphere soft component, and gamma = 2.7 +- 0.4 for the blackbody soft component. Timing analysis shows three peaks in the pulse profile, separated by about 0.3 in phase. Energy-resolved timing provides evidence for pulse profile variation with energy. The higher energy (E > 1.8 keV) profile shows significantly higher pulsed fraction., Comment: 4 pages, 2 figures, To appear in "Neutron Stars in Supernova Remnants" (ASP Conference Proceedings), eds P. O. Slane and B. M. Gaensler Corrected TYPOS

We present the Chandra LETGS X-ray spectrum of the nearby (~60 pc) neutron star RX J1856.5-3754. Detailed spectral analysis of the combined X-ray and optical data rules out the nonmagnetic neutron star atmosphere models with hydrogen, helium, iron and solar compositions. We also conclude that strongly magnetized atmosphere models are unable to represent the data. The data can be explained with a two-component blackbody model. The harder component with temperature of kT_bb~63 eV and a radius R_bb~2.2 km of the emitting region well fits the X-ray data and can be interpreted as radiation from a hot region on the star's surface., Comment: 4 pages, 3 color figures; acceped by A&A Letters; http://www.xray.mpe.mpg.de/~burwitz/burwitz_refereed.html

We report the results of the spectral analysis of two observations of the Vela pulsar with the Chandra X-ray observatory. The spectrum of the pulsar does not show statistically significant spectral lines in the observed 0.25-8.0 keV band. Similar to middle-aged pulsars with detected thermal emission, the spectrum consists of two distinct components. The softer component can be modeled as a magnetic hydrogen atmosphere spectrum - for the pulsar magnetic field $B=3\times 10^{12}$ G and neutron star mass $M=1.4 M_\odot$ and radius $R^\infty =13$ km, we obtain $\tef^\infty =0.68\pm 0.03$ MK, $L_{\rm bol}^\infty = (2.6\pm 0.2)\times 10^{32}$ erg s$^{-1}$, $d=210\pm 20$ pc (the effective temperature, bolometric luminosity, and radius are as measured by a distant observer). The effective temperature is lower than that predicted by standard neutron star cooling models. A standard blackbody fit gives $T^\infty =1.49\pm 0.04$ MK, $L_{\rm bol}^\infty=(1.5\pm 0.4)\times 10^{32} d_{250}^2$ erg s$^{-1}$ ($d_{250}$ is the distance in units of 250 pc); the blackbody temperature corresponds to a radius, $R^\infty =(2.1\pm 0.2) d_{250}$ km, much smaller than realistic neutron star radii. The harder component can be modeled as a power-law spectrum, with parameters depending on the model adopted for the soft component - $\gamma=1.5\pm 0.3$, $L_x=(1.5\pm 0.4)\times 10^{31} d_{250}^2$ erg s$^{-1}$ and $\gamma=2.7\pm 0.4$, $L_x=(4.2\pm 0.6)\times 10^{31} d_{250}^2$ erg s$^{-1}$ for the hydrogen atmosphere and blackbody soft component, respectively ($\gamma$ is the photon index, $L_x$ is the luminosity in the 0.2--8 keV band). The extrapolation of the power-law component of the former fit towards lower energies matches the optical flux at $\gamma\simeq 1.35$--1.45., Comment: Submitted to ApJ, three figures; color figure 1 can be found at http://www.xray.mpe.mpg.de/~zavlin/pub_list.html

We present the XMM-Newton first light image, taken in January 2000 with the EPIC pn camera during the instrument's commissioning phase, when XMM-Newton was pointing towards the Large Magellanic Cloud (LMC). The field is rich in different kinds of X-ray sources: point sources, supernova remnants (SNRs) and diffuse X-ray emission from LMC interstellar gas. The observations are of unprecedented sensitivity, reaching a few 10^32 erg/s for point sources in the LMC. We describe how these data sets were analysed and discuss some of the spectroscopic results. For the SNR N157B the power law spectrum is clearly steeper than previously determined from ROSAT and ASCA data. The existence of a significant thermal component is evident and suggests that N157B is not a Crab-like but a composite SNR. Most puzzling is the spectrum of the LMC hot interstellar medium, which indicates a significant overabundance of Ne and Mg of a few times solar., Comment: to be published in A&A 365

The central source of the supernova remnant PKS 1209-52 was observed with the Advanced CCD Imaging Spectrometer aboard Chandra X-ray observatory on 2000 January 6-7. The use of the Continuos Clocking mode allowed us to perform the timing analysis of the data with time resolution of 2.85 ms and to find a period P=0.42412927+/-2.3e-7 s. The detection of this short period proves that the source is a neutron star. It may be either an active pulsar with unfavorably directed radio beam or a truly radio-silent neutron star whose X-ray pulsations are caused by a nonuniform distribution of surface temperature. To infer the actual properties of this neutron star, the period derivative should be measured., Comment: submitted to ApJ Letters

The central pointlike X-ray source of the Cas A supernova remnant was discovered in the Chandra First Light Observation and found later in the archival ROSAT and Einstein images. The analysis of these data does not show statistically significant variability of the source. The power-law fit yields the photon index 2.6-4.1, and luminosity (2-60)e34 erg/s, for d=3.4 kpc. The power-law index is higher, and the luminosity lower, than those observed fromvery young pulsars. One can fit the spectrum equally well with a blackbody model with T=6-8 MK, R=0.2-0.5 km, L=(1.4-1.9)e33 erg/s. The inferred radii are too small, and the temperatures too high, for the radiationcould be interpreted as emitted from the whole surface of a uniformly heated neutron star. Fits with the neutron star atmosphere models increase the radius and reduce the temperature, but these parameters are still substantially different from those expected for a young neutron star. One cannot exclude, however, that the observed emission originates from hot spots on a cooler neutron star surface. Because of strong interstellar absorption, the possible low-temperature component gives a small contribution to the observed spectrum; an upper limit on the (gravitationally redshifted) surface temperature is < 1.9-2.3 MK. Amongst several possible interpretations, we favor a model of a strongly magnetized neutron star with magnetically confined hydrogen or helium polar caps on a cooler iron surface. Alternatively, the observed radiation may be interpreted as emitted by a compact object (more likely, a black hole) accreting from a fossil disk or from a late-type dwarf in a close binary., Comment: 12 pages, 2 figures, submitted to ApJL

Since the opacity of a magnetized plasma depends on polarization of radiation, the radiation emergent from atmospheres of neutron stars with strong magnetic fields is expected to be strongly polarized. The degree of linear polarization, typically ~10-30%, depends on photon energy, effective temperature and magnetic field. The spectrum of polarization is more sensitive to the magnetic field than the spectrum of intensity. Both the degree of polarization and the position angle vary with the neutron star rotation period so that the shape of polarization pulse profiles depends on the orientation of the rotational and magnetic axes. Moreover, as the polarization is substantially modified by the general relativistic effects, observations of polarization of X-ray radiation from isolated neutron stars provide a new method for evaluating the mass-to-radius ratio of these objects, which is particularly important for elucidating the properties of the superdense matter in the neutron star interiors., Comment: 7 figures, to be published in ApJ

There are several supernova remnants which contain unresolved X-ray sources close to their centers, presumably radio-quiet neutron stars. To prove that these objects are indeed neutron stars, to understand the origin of their X-ray radiation, and to explain why they are radio-quiet, one should know their periods and period derivatives. We searched for pulsations of the X-ray flux from the radio-quiet neutron star candidate RX J0822-4300 near the center of the Puppis A supernova remnant observed with the ROSAT PSPC and HRI. A standard timing analysis of the separate PSPC and HRI data sets does not allow one to detect the periodicity unequivocally. However, a thorough analysis of the two observations separated by 4.56 yr enabled us to find a statistically significant period $P\simeq 75.3$ ms and its derivative $\dot{P}\simeq 1.49\times 10^{-13}$ s s$^{-1}$. The corresponding characteristic parameters of the neutron star, age $\tau=P/(2\dot{P})=8.0$ kyr, magnetic field $B=3.4\times 10^{12}$ G, and rotational energy loss $\dot{E}=1.4\times 10^{37}$ erg s$^{-1}$, are typical for young radio pulsars. Since the X-ray radiation has a thermal-like spectrum, its pulsations may be due to a nonuniform temperature distribution over the neutron star surface caused by anisotropy of the heat conduction in the strongly magnetized crust., Comment: 9 pages, 2 postscript figures, to appear in ApJ Letters; an acknowledgment is added

We re-analyze the available X-ray spectral data of the type I bursting neutron star transients Aql X-1, Cen X-4, and 4U 1608-522 using realistic hydrogen atmosphere models. Previous spectral fits assumed a blackbody spectrum; because the free-free dominated photospheric opacity decreases with increasing frequency, blackbody spectral fits overestimate the effective temperature and underestimate, by as much as two orders of magnitude, the emitting area. Hydrogen atmosphere spectral models, when fit to the available observational data, imply systematically larger emission area radii, consistent with the canonical 10 km radius of a neutron star. This suggests that a substantial fraction of the quiescent luminosity is thermal emission from the surface of the neutron star. The magnitude of the equivalent hydrogen column density toward these systems, however, presents a considerable systematic uncertainty, which can only be eliminated by high signal-to-noise X-ray spectral measurements (e.g., with AXAF or XMM) which would permit simultaneous determination of the equivalent hydrogen column density, emission area, and thermal temperature., Comment: 16 Pages, LaTeX, no Figures, accepted by ApJ

We re-analyzed soft X-ray data collected with the ROSAT and ASCA observatories on a candidate neutron star (NS) near the center of the supernova remnant PKS 1209-52. We fitted the observed spectra with NS atmosphere models. The hydrogen atmosphere fits yield more realistic parameters of the NS and the intervening hydrogen column than the traditional blackbody fit. In particular, for a NS of mass 1.4 M_\odot and radius 10 km, we obtained the NS surface temperature T_{eff}=(1.4-1.9)*10^6 K and distance d=1.6-3.3 kpc versus T=(4.2-4.6)*10^6 K and (implausible) d=11-13 kpc for the blackbody fit, at a 90% confidence level. Our fits suggest that the surface magnetic field is either very weak, B < 10^{10} G, or it exceeds 2*10^{12} G. The hydrogen column density inferred from the atmosphere fits, n_H=(0.7-2.2)*10^{21} cm^{-2}, agrees fairly well with independent estimates obtained from UV observations of nearby stars, radio data, and X-ray spectrum of the shell of the supernova remnant, whereas the blackbody and power-law fits give considerably lower and greater values, n_H=(0.2-0.4)*10^{21} and (5.2-7.0)*10^{21} cm^{-2}, respectively. The inferred NS surface temperature is consistent with standard NS cooling models., Comment: 4 figures, l-aa.sty; submitted to A&A

Properties of X-ray radiation emitted from the polar caps of a radio pulsar depend not only on the cap temperature, size, and position, but also on the surface chemical composition, magnetic field, and neutron star's mass and radius. Fitting the spectra and the light curves with neutron star atmosphere models enables one to infer these parameters. As an example, we present here results obtained from the analysis of the pulsed X-ray radiation of a nearby millisecond pulsar J0437-4715. In particular, we show that stringent constraints on the mass-to-radius ratio can be obtained if orientations of the magnetic and rotation axes are known, e.g., from the radio polarization data., Comment: 2 figures, aasms4.sty; accepted for publication in ApJLetters

We show that the soft X-ray spectra and light curves observed with the ROSAT and EUVE from the closest known millisecond pulsar J0437--4715 can be interpreted as thermal radiation from two hot polar caps whose emitting layers (atmospheres) are comprised of hydrogen. The simplest model yields a uniform temperature of $(0.8-0.9)\times 10^6 K$ within a cap radius of $0.7-0.9 km$. The spectral fits indicate that the temperature may be nonuniformly distributed along the cap surface. The distribution can be approximated by a central core heated up to $(1-2)\times 10^6 K$ within a radius of $0.2-0.4 km$, surrounded by a colder rim with temperatures $(3-5)\times 10^5 K$ extending out to $2-6 km$. The polar cap interpretation implies low column densities, $(1-3)\times 10^{19} cm^{-2}$, and a high degree of ionization, $> 20$%, of the interstellar hydrogen towards the pulsar. The inferred bolometric luminosity of the polar caps, $(1.0-1.6)\times 10^{30} erg s^{-1}$, is in excellent agreement with the predictions of the slot-gap model of radio pulsars developed by Arons and his coworkers. Similar polar cap radiation should be emitted by other millisecond pulsars, although in some of them (e.~g., PSR B1821--24) the soft X-ray flux is dominated by the nonthermal radiation from pulsar magnetospheres., Comment: 17 pages with 11 figures l-aa.sty, accepted for publication in Astronomy and Astrophysics

The properties of X-ray radiation from the polar caps predicted by the radio pulsar models depend on the surface chemical composition, magnetic field and star's mass and radius as well as on the cap temperature, size and position. Fitting the radiation spectra and light curves with the neutron star atmosphere models enables one to infer these parameters. We present here results obtained from the analysis of the soft X-ray radiation of PSR J0437-4715. In particular, with the aid of radio polarization data, we put constraints on the pulsar mass-to-radius ratio., Comment: one author name deleted. 3 pages, including 1 figure, 'sprocl.sty' macros (attached), to be published in Proceedings of 18th Texas Symposium on Relativistic Astrophysics, A. Olinto, J. Frieman and D. Schramm (eds.), World Sci. Press

We show that an analysis of multiwavelength observations of isolated neutron stars based on neutron star atmosphere models can be used not only to evaluate the neutron star effective temperature, but also to determine chemical composition of its surface. To demonstrate how this new method can be applied to a specific object, we chose the old isolated neutron star candidate RXJ1856.5-3754, whose soft X-ray radiation has been studied recently by Walter et al. (1996). We fitted the soft X-ray spectrum of this object with neutron star atmosphere models of different chemical compositions and used these fits to calculate the source spectrum over a broad wavelength range. We showed, in particular, that the optical/UV flux expected from this object depends drastically on the composition of its surface. In particular, the neutron star covered with hydrogen would be 5-6 magnitudes brighter than the neutron star with an iron surface. The object should also be observable with $EUVE$; the EUV flux is expected to be almost twice higher for the iron surface than for the hydrogen one. Thus, multiwavelength observations of this object would enable one to examine, for the first time, chemical composition of the neutron star surface. The method proposed can be applied to other nearby isolated neutron stars., Comment: a few misprints corrected, aaspp4.sty, accepted for publication in ApJ Letters

We present a detailed investigation of neutron star atmospheres with low magnetic fields, $B < 10^8 -10^{10}$ G, which do not affect opacities and equation of state of the atmospheric matter. We compute the atmospheric structure, emergent spectral fluxes and specific intensities for hydrogen, helium and iron atmospheres in a wide domain of effective temperatures and gravitational accelerations expected for neutron stars. The iron atmospheres are computed with the opacities and equations of state from the OPAL opacity library. We show that the model atmosphere spectra are substantially different from the blackbody spectra. For light element atmospheres, the flux is greater than the blackbody flux, and the spectrum is harder, at high photon energies, whereas at low energies the spectral flux follows the Rayleigh-Jeans law with a (surface) temperature lower than the effective temperature. The spectra of iron atmospheres display prominent spectral features in the soft X-ray range. The emergent specific intensity is anisotropic, with the anisotropy depending on energy. These properties of the atmospheric radiation should be taken into account for the proper interpretation of the thermal component of the neutron star radiation detectable at X-ray through UV energies. In particular, fitting of the thermal component with the blackbody model may result in substantially incorrect parameters of the radiating regions. We discuss an application of our atmosphere models to the nearby millisecond pulsar PSR J0437-4715., Comment: 13 pages in 'l-aa.sty' with 9 figures; A&A 1996, accepted

We use models of magnetized hydrogen atmospheres to fit the soft X-ray thermal spectrum of the Vela pulsar obtained by ROSAT. The distance and hydrogen column density required by our fits are in good agreement with other independents estimates and can be considered as an argument for the presence of hydrogen at the surface of this pulsar. The low temperature obtained, T = 7.85 +/- 0.25 10^5 K, is below the predictions of the `standard' model of neutron star cooling and strengthens previous claims for the necessity of the occurrence of fast neutrino cooling in this neutron star., Comment: 2 pages Latex with 1 figure and 1 heading in postscript. Style file: xray-wuerzburg-95.sty

We present first results of modeling convective photospheres of neutron stars. We show that in photospheres composed of the light elements convection arises only at relatively low effective temperatures (< 3-5*10^4 K), whereas in the case of iron compositon it arises at < 3*10^ K. Convection changes the depth dependence of the photosphere temperature and the shapes of the emergent spectra. Thus, it should be taken into account for the proper interpretation of EUV/soft-X-ray observations of the thermal radiation from neutron stars., Comment: 2 pages, from Wuerzburg X-ray Conference (Sep. 25-29, 1995); final version; ask for psfile to: zavlin@rosat.mpe-garching.mpg.de

We present a model to explain the decrease in the amplitude of the pulse profile with increasing energy observed in Geminga's soft X-ray surface thermal emission. We assume the presence of plates surrounded by a surface with very distinct physical properties: these two regions emit spectra of very distinct shapes which present a crossover, the warm plates emitting a softer spectrum than the colder surrounding surface. The strongly pulsed emission from the plates dominates at low energy while the surroundings emission dominates at high energy, producing naturally a strong decrease in the pulsed fraction. In our illustrative example the plates are assumed to be magnetized while the rest of the surface is field free. This plate structure may be seen as a schematic representation of a continuous but very nonuniform distribution of the surface magnetic field or as a quasi realistic structure induced by past tectonic activity on Geminga., Comment: 10 pages, AASTeX latex, + 3 figures (compressed 7 uuencoded). Submitted to Ap. J. Lett

We investigate the anisotropy of the thermal radiation emitted by a surface element of a neutron star atmosphere (e.g., by a polar cap of a radio pulsar). Angular dependences of the partial fluxes at various photon energies, and spectra at various angles are obtained for different values of the effective temperature T(sub eff) and magnetic field strength B, and for different directions of the magnetic field. It is shown that the local radiation of the magnetized neutron star atmospheres is highly anisotropic, with the maximum flux emitted in the magnetic field direction. At high B the angular dependences in the soft X-ray range have two maxima, a high narrow peak along B and a lower and broader maximum at intermediate angles. The radiation is strongly polarized, the modulation of the degree of polarization due to the rotation of the neurtron star may be much higher than that for the radiative flux. The results obtained are compared with recent ROSAT observations of the thermal-like radiation from the radio pulsars PSR 1929+10 and PSR J0437-4715.

Model neutron star atmospheres are calculated for typical cooling stars with a strong magnetic field and effective temperatures of 10 exp 5 to 10 exp 6 K. The effect of anisotropic photon diffusion in two normal modes are examined under the assumption that the opacity is due solely to the bremsstrahlung and Thomson scattering processes under conditions of LTE that are expected to prevail at the temperatures and densities obtained. The main aspects of anisotropic photon diffusion, and an original procedure for calculating model atmospheres and emitted spectra are discussed. Representative calculated spectra are given, and it is found that the hard spectral excess characterizing the nonmagnetic case, while still present, becomes less prominent in the presence of magnetic fields in the range of 10 exp 11 to 10 exp 13 G.