Table of contents

Nonspherical collapse is modeled, under the Zeldovich approximation, by six-dimensional random walks of the initial deformation tensor field. The collapse boundary adopted here is a slightly modified version of that proposed by Chiueh and Lee. Not only does the mass function agree with the fitting formula of Sheth and Tormen, but the bias function and conditional mass function constructed by this model are also found to agree reasonably well with the simulation results of Jing and Somerville et al., respectively. In particular, by introducing a small mass gap, we find a fitting formula for the conditional mass function, which works well even at small time intervals between parent and progenitor halos during the merging history.

We present a detailed nonspherical modeling of dark matter halos on the basis of a combined analysis of high-resolution halo simulations (12 halos with N ~ 10⁶ particles within their virial radius) and large cosmological simulations (five realizations with N = 512³ particles in a 100 h^-1 Mpc box size). The density profiles of those simulated halos are well approximated by a sequence of the concentric triaxial distribution with their axis directions being fairly aligned. We characterize the triaxial model quantitatively by generalizing the universal density profile, that has previously been discussed only in the framework of the spherical model. We obtain a series of practically useful fitting formulae in applying the triaxial model: the mass and redshift dependence of the axis ratio, the mean of the concentration parameter, and the probability distribution functions of the axis ratio and the concentration parameter. These accurate fitting formulae form a complete description of the triaxial density profiles of halos in cold dark matter models. Our current description of the dark halos will be particularly useful in predicting a variety of nonsphericity effects, to a reasonably reliable degree, including the weak and strong lens statistics, the orbital evolution of galactic satellites and triaxiality of galactic halos, and the nonlinear clustering of dark matter. In addition, this provides a useful framework for the nonspherical modeling of the intracluster gas, which is crucial in discussing the gas and temperature profiles of X-ray clusters and the Hubble constant estimated via the Sunyaev-Zeldovich effect.

Using 220 gamma-ray burst (GRB) redshifts and luminosities derived from the luminosity-variability relationship of Fenimore & Ramirez-Ruiz, we show that there exists a significant correlation between the GRB luminosity and redshift. In particular, we find that the evolution of the average luminosity can be parameterized as L ∝ (1 + z)^1.4±~0.5, where z is the burst redshift. We discuss the possible reasons behind this evolution and compare it with that of other known sources that exhibit similar behavior. In addition, we use nonparametric statistical techniques to independently estimate the distributions of the luminosity and redshift of bursts, accounting for the evolution (in contrast to previous studies, which have assumed that the luminosity function is independent of redshift). We present these distributions and discuss their implications. Most significantly, we find a comoving rate density of GRBs that continues to increase to (1 + z) ≳ 10. From this estimate of the GRB rate density, we then use the population synthesis codes of Fryer et al. to estimate the star formation rate at high redshifts, for different progenitor models of GRBs. We find that no matter what the progenitor or population synthesis model, the star formation rate increases or remains constant to very high redshifts (z ≳ 10).

Weak gravitational lensing by the intervening large-scale structure of the universe induces high-order correlations in the cosmic microwave background temperature and polarization fields. We construct minimum variance estimators of the intervening mass distribution out of the six quadratic combinations of the temperature and polarization fields. Polarization begins to assist in the reconstruction when E-mode mapping becomes possible on degree-scale fields, i.e., for an experiment with a noise level of ~40 μK arcmin and beam of ~7', similar to the Planck experiment; surpasses the temperature reconstruction at ~26 μK arcmin and 4'; and yet continues to improve the reconstruction until the lensing B-modes are mapped to l ~ 2000 at ~0.3 μK arcmin and 3'. Ultimately, the correlation between the E- and B-modes can provide a high signal-to-noise ratio mass map out to multipoles of L ~ 1000, extending the range of temperature-based estimators by nearly an order of magnitude. We outline four applications of mass reconstruction: measurement of the linear power spectrum in projection to the cosmic variance limit out to L ~ 1000 (or wavenumbers 0.002 ≲ k ≲ 0.2 in h Mpc^-1), cross-correlation with cosmic shear surveys to probe the evolution of structure tomographically, cross-correlation of the mass and temperature maps to probe the dark energy, and the separation of lensing and gravitational wave B-modes.

Recently, it has been found that the field traced by QSOs' Lyα forests is intermittent on small scales. Intermittent behavior is essential for understanding the statistics and dynamics of cosmic gravitational clustering in the nonlinear regime. The most effective method of describing intermittency uses the structure functions and the intermittent exponent, which measure the scale and order dependencies of the ratio between the higher order and second-order moments of the field. These properties can be used not only to confirm the non-Gaussianity of fields but also to detect the type of non-Gaussianity. In this paper, we calculate the structure function and intermittent exponent of (1) Keck data, which consists of 28 high-resolution, high signal-to-noise ratio QSO Lyα absorption spectra, and (2) Lyα forest simulation samples produced via the pseudohydrodynamic scheme for the low-density cold dark matter model and warm dark matter (WDM) model with particle mass m_W = 300, 600, 800, and 1000 eV. Aside from the WDM model with m_W = 300 eV, the simulation samples are in agreement with observations in the context of the power spectrum. We find, however, that the intermittent behavior of all the simulation samples is substantially inconsistent, both quantitatively and qualitatively, with the Keck data. Specifically, (1) the structure functions of the simulation samples are significantly larger than that of Keck data on scales k ≥ 0.1 km^-1 s; (2) the intermittent exponents of the simulation samples are more negative than that of Keck data on all redshifts considered; and (3) the order dependence of the structure functions of simulation samples is closer to the intermittency of hierarchical clustering on all scales, while the Keck data are closer to a lognormal field on small scales. These differences are independent of noise and show that the intermittent evolution modeled by the pseudohydrodynamic simulation is substantially different from observations, even though they are in good agreement with each other in terms of second- and lower order statistics. This result also shows that "weakly" clustered samples, such as the high-resolution Lyα absorption spectrum, are effective in testing dynamical models of structure formation if their intermittent features are considered.

Supernova-driven outflows from early galaxies may have had a large impact on the kinetic and chemical properties of the intergalactic medium (IGM). We use three-dimensional Monte Carlo cosmological realizations of a simple linear peaks model to track the time evolution of such metal-enriched outflows and their feedback on galaxy formation. We find that at most 30% of the IGM by volume is enriched to values above 10^-3Z_☉ in models that include only objects that cool by atomic transitions. The majority of enrichment occurs relatively early (5 ≲ z ≲ 12) and leads to a mass-averaged cosmological metallicity between 10^-3 and 10^-1.5Z_☉. The inclusion of Population III objects that cool through H₂ line emission has only a minor impact on these results: increasing the mean metallicity and filling factor by at most a factor of 1.4 and moving the dawn of the enrichment epoch to z ≈ 14 at the earliest. Thus, enrichment by outflowing galaxies is likely to have been incomplete and inhomogeneous, biased to the areas near the starbursting galaxies themselves. Models with a 10% star formation efficiency can satisfactorily reproduce the nearly constant (2 ≤ z ≤ 5, Z ≈ 3.5 × 10^-4Z_☉) metallicity of the low column density Lyα forest derived by Songaila in 2001, an effect of the decreasing efficiency of metal loss from larger galaxies. Finally, we show that IGM enrichment is intimately tied to the ram-pressure stripping of baryons from neighboring perturbations. This results in the suppression of at least 20% of the dwarf galaxies in the mass range ~3 × 10⁸-3 × 10⁹M_☉ in all models with filling factors greater than 2% and an overall suppression of ~50% of dwarf galaxies in the most observationally favored model.

I present the results of a search for intergalactic hydrogen clouds in voids. Clouds are detected by their H I Lyα absorption lines in the Hubble Space Telescope spectra of low-redshift active galactic nuclei. The parameter with which the environments of clouds are characterized is the tidal field, for this places a lower limit on the cloud mass density that is dynamically stable against disruption. The summing of the tidal fields along these lines of sight is managed by the use of galaxy redshift catalogs. The analytical methodology employed here has been designed to detect gas clouds whose expansion following reionization is restrained by dark matter perturbations. The end products of the analysis of data are the cloud equivalent width distribution functions (EWDFs) of catalogs formed by sorting clouds according to various tidal field upper or lower limits. Cloud EWDFs are steep in voids [d log(d/dz)/d log() = ~ -1.5 ± 0.2] but flatter in high tidal field zones ( ~ -1.5 ± 0.6). The most probable cloud Doppler parameters are ~30 km s^-1 in voids and ~60 km s^-1 in proximity to galaxies. In voids the cumulative line density at low EW ( ≃ 15 mÅ) is essentially equal to that of the mean EWDF, ~500 per unit redshift. The void filling factor is found to be 0.87 ≤ f_v ≤ 0.94. The void EWDF is remarkably uniform over this volume, with a possible tendency for more massive clouds to be in void centers. The size and nature of the void cloud population suggested by this study is completely unanticipated by the results of published three-dimensional simulations, which predict that most clouds are in filamentary structures around galaxy concentrations and that very few observable absorbers would lie in voids. Strategies for modeling this population are briefly discussed.

We present spectroscopic observations with the Keck Low-Resolution Imaging Spectrometer of LBQS 0015+0239, a pair of quasars at z = 2.45 with a separation of Δθ = 22 (projected linear distance of 17.8 h kpc, for Ω_m = 0.3, Λ = 0.7). Lensing is an unlikely interpretation for the images since the spectra show significant differences in the N V and C IV emission-line profiles and there is no luminous galaxy at the anticipated lens position. Rather, we interpret this pair as the highest redshift known example of a binary quasar. The redshift difference of 661 ± 173 km s^-1 between the two components is inconsistent with the lensing expectation but is consistent with the line-of-sight velocity difference of a bound pair of galaxies. The spectra show associated narrow metal absorption in highly ionized species of C and N, with a systemic velocity difference of 492 ± 6 km s^-1. It is likely that this absorption arises in a group or cluster of galaxies surrounding the quasar pair. After a thorough search, the Large Bright Quasar Survey is known to contain one gravitational lens and four probable binary pairs. The existence of close binaries is evidence that quasar activity can be triggered by galaxy interactions. We speculate that the close pair LBQS 0015+0239 is a rare example of the premerger state of two supermassive black holes.

The bright blazar Mrk 421 has been observed four times for uninterrupted durations of ~9-13 hr during the performance verification and calibration phases of the XMM-Newton mission. The source was strongly variable in all epochs, with variability amplitudes that generally increased to higher energy bands. Although the detailed relationship between soft (0.1-0.75 keV) and hard (2-10 keV) bands differed from one epoch to the next, in no case was there any evidence for a measurable interband lag, with robust upper limits of |τ| < 0.08 hr in the best-correlated light curves. This is in conflict with previous claims of both hard and soft lags of ~1 hr in this and other blazars. However, previous observations suffered a repeated 1.6 hr feature induced by the low-Earth orbital period, a feature that is not present in the uninterrupted XMM-Newton data. The new upper limit on |τ| leads to a lower limit on the magnetic field strength and Doppler factor of Bδ^1/3 ≳ 4.7 G, mildly out of line with the predictions from a variety of homogeneous synchrotron self-Compton emission models in the literature of Bδ^1/3 = 0.2-0.8 G. Time-dependent spectral fitting was performed on all epochs, and no detectable spectral hysteresis was seen. We note, however, that the source exhibited significantly different spectral evolutionary behavior from one epoch to the next, with the strongest correlations in the first and last and an actual divergence between soft and hard X-ray bands in the third. This indicates that the range of spectral variability behavior in Mrk 421 is not fully described in these short snippets; significantly longer uninterrupted light curves are required and can be obtained with XMM-Newton.

We present results from a 900 ks exposure of NGC 3783 with the High-Energy Transmission Grating Spectrometer on board the Chandra X-Ray Observatory. The resulting X-ray spectrum, which covers the 0.5-10 keV energy range, has the best combination of signal-to-noise ratio and resolution ever obtained for an AGN. This spectrum reveals absorption lines from H-like and He-like ions of N, O, Ne, Mg, Al, Si, and S. There are also possible absorption lines from H-like and He-like Ar and Ca as well as H-like C. We also identify inner-shell absorption from lower ionization ions such as Si VII-Si XII and S XII-S XIV. The iron absorption spectrum is very rich; L-shell lines of Fe XVII-Fe XXIV are detected, as well as probable resonance lines from Fe XXV. A strong complex of M-shell lines from iron ions is also detected in the spectrum. The absorption lines are blueshifted relative to the systemic velocity by a mean velocity of -590 ± 150 km s^-1. We resolve many of the absorption lines, and their mean FWHM is 820 ± 280 km s^-1. We do not find correlations between the velocity shifts or the FWHMs with the ionization potentials of the ions. Most absorption lines show asymmetry, having more extended blue wings than red wings. In O VII we have resolved this asymmetry to be from an additional absorption system at approximately -1300 km s^-1. The two X-ray absorption systems are consistent in velocity shift and FWHM with the ones identified in the UV lines of C IV, N V, and H I. Equivalent width measurements for all absorption and emission lines are given and column densities are calculated for several ions. We resolve the narrow Fe Kα line at 6398.2 ± 3.3 eV to have an FWHM of 1720 ± 360 km s^-1, which suggests that this narrow line may be emitted from the outer part of the broad-line region or the inner part of the torus. We also detect a "Compton shoulder" redward of the narrow Fe Kα line, which indicates that it arises in cold, Compton-thick gas.

We present deep Chandra spectral imaging of the dwarf starburst galaxy NGC 1569. The unprecedented spatial resolution allows us to spatially identify the components of the integrated X-ray spectrum. Fitted spectral models require an intrinsic absorption component and higher metal abundances than previous studies indicated. Our results provide the first direct evidence for metal-enriched winds from dwarf starburst galaxies. We identify 14 X-ray point sources in NGC 1569. Most have properties consistent with those of high-mass X-ray binaries, but one is a steep-spectrum radio source that is probably a supernova remnant. The X-ray luminosity of NGC 1569 is dominated by diffuse, thermal emission from the disk (0.7 keV) and bipolar halo (0.3 keV). Photoelectric absorption from the inclined H I disk hardens the X-ray spectrum on the northern side of the disk relative to the southern side. Requiring the fitted absorption column to match the H I column measured at 21 cm implies that the metallicity of the H I disk is significantly less than solar but greater than 0.1 Z_☉. Hence, much of the H I is enriched to levels comparable to the metallicity of the H II regions [O/H = 0.2(O/H)_☉]. The X-ray color variations in the halo are inconsistent with a free-streaming wind and probably reveal the location of shocks created by the interaction of the wind with a gaseous halo. The X-ray spectrum of the diffuse gas presents strong emission lines from α-process elements. Fitted models require α-element abundances greater than 0.25 Z_{α, ☉} and ratios of α-elements to iron 2-4 times higher than the solar ratio. The best fit to the spectrum is obtained with solar mass fractions for the α-elements, 1.0 Z_{α, ☉}, but a degeneracy between the metallicity and the spectral normalization prevents us from deriving an upper limit on the wind metallicity from the X-ray spectrum alone. We argue, however, that abundances larger than 2.0 Z_{α, ☉} pose awkward implications for the dynamical evolution of the wind based on our knowledge of the starburst properties. For consistency with our best-fitting abundances, the mass of interstellar gas entrained in the wind must be about 9 times the mass of stellar ejecta in the wind. Most of the oxygen carried by the wind comes from the stellar ejecta rather than entrained interstellar gas. The estimated mass of oxygen in the hot wind, 34,000 M_☉, is similar to the oxygen yield of the current starburst. Apparently the wind carries nearly all the metals ejected by the starburst. These metals appear destined to contribute to the enrichment of the intergalactic medium. Much of the nucleosynthesis in NGC 1569 must have occurred during less violent periods of star formation, however, because our measurements imply that the neutral gas disk holds at least 5 times more oxygen than wind.

We discuss the nature of the unidentified EGRET source 3EG J1621+8203. In an effort to identify the gamma-ray source, we have examined X-ray images of the field from ROSAT PSPC, ROSAT HRI, and ASCA GIS. Of the several faint X-ray point sources in the error circle of 3EG J1621+8203, most are stars or faint radio sources, unlikely to be counterparts to the EGRET source. The most notable object in the gamma-ray error box is the bright FR I radio galaxy NGC 6251. If 3EG J1621+8203 corresponds to NGC 6251, then it would be the second radio galaxy to be detected in high-energy gamma rays after Cen A, which provided the first clear evidence of the detection above 100 MeV of an active galactic nucleus (AGN) with a large-inclination jet. If the detection of more radio galaxies by EGRET has been limited by its threshold sensitivity, there exists the exciting possibility that new high-energy gamma-ray instruments, with much higher sensitivity, will detect a larger number of radio galaxies in the future.

We have imaged the H92α and H75α radio recombination line (RRL) emissions from the starburst galaxy NGC 253 with a resolution of ~4 pc. The peak of the RRL emission at both frequencies coincides with the unresolved radio nucleus. Both lines observed toward the nucleus are extremely wide, with FWHMs of ~200 km s^-1. Modeling the RRL and radio continuum data for the radio nucleus shows that the lines arise in gas whose density is ~10⁴ cm^-3 and mass is a few thousand M_☉, which requires an ionizing flux of (6-20) × 10⁵¹ photons s^-1. We consider a supernova remnant (SNR) expanding in a dense medium, a star cluster, and also an active galactic nucleus (AGN) as potential ionizing sources. Based on dynamical arguments, we rule out an SNR as a viable ionizing source. A star cluster model is considered, and the dynamics of the ionized gas in a stellar-wind driven structure are investigated. Such a model is only consistent with the properties of the ionized gas for a cluster younger than ~10⁵ yr. The existence of such a young cluster at the nucleus seems improbable. The third model assumes the ionizing source to be an AGN at the nucleus. In this model, it is shown that the observed X-ray flux is too weak to account for the required ionizing photon flux. However, the ionization requirement can be explained if the accretion disk is assumed to have a big blue bump in its spectrum. Hence, we favor an AGN at the nucleus as the source responsible for ionizing the observed RRLs. A hybrid model consisting of an inner advection-dominated accretion flow disk and an outer thin disk is suggested, which could explain the radio, UV, and X-ray luminosities of the nucleus.

We study the far-infrared emission properties of the nearby starburst galaxy NGC 253 based on IRAS maps and an ISOPHOT map at 180 μm. Based on the analysis of the light profiles, we have been able to identify three main structural components: an unresolved nuclear component, an exponential disk, and a kiloparsec-scale bar. In addition, we also found a ring structure at the end of the bar that is particularly conspicuous at 12 μm. The spectral energy distribution of each morphological component has been modeled as thermal dust emission at different temperatures. The unresolved nuclear component is dominated by cold dust emission (T ≃ 50 K), whereas the disk emission is dominated by very cold dust (T ≃ 16 K) plus a contribution from cold dust (T ≃ 55 K). The bar emission corresponds mainly to cold dust (T ≃ 23 K) plus a warm component (T ≃ 148 K). We detect an extension of the disk emission due to very cold dust, which contributes a large fraction (94%) of the total dust mass of the galaxy. The estimated total dust mass is 8.2 × 10⁷M_☉.

We determine the extinction curve in the z_l = 0.83 lens galaxy of the gravitational lens SBS 0909+532 from the wavelength dependence of the flux ratio between the lensed quasar images (z_s = 1.38) from 3400 to 9200 Å. It is the first measurement of an extinction curve at a cosmological distance of comparable quality to those obtained within the Galaxy. The extinction curve has a strong 2175 Å feature, a noteworthy fact because it has been weak or nonexistent in most estimates of extinction curves outside the Galaxy. The extinction curve is fitted well by a standard R_V = 2.1 ± 0.9 Galactic extinction curve. If we assume standard Galactic extinction laws, the estimated dust redshift of z = 0.88 ± 0.02 is in good agreement with the spectroscopic redshift of the lens galaxy. The widespread assumption that SMC extinction curves are more appropriate models for cosmological dust may be incorrect.

We present the results of our search for faint Local Group dwarf galaxies in compact high-velocity clouds (HVCs). We used digitized Palomar Observatory Sky Survey (POSS) data to examine 1 deg² of sky around each of ~250 northern hemisphere HVCs. The POSS images were processed to remove foreground stars and large-scale backgrounds, smoothed to enhance arcminute-sized low surface brightness features, and then compared with the original plates. Using this technique, we located 60 candidate dwarf galaxies in the ~250 deg² that we surveyed. Follow-up observations of these candidates have revealed several distant clusters of galaxies and a number of possible Galactic cirrus clouds, but no Local Group dwarfs. It appears that many of the low surface brightness features in the sky survey data are plate flaws. The second-generation red POSS plates are sensitive down to surface brightness levels of 25-26 mag arcsec^-2, and our follow-up images reached 10 σ limiting magnitudes of R = 21-23 for point sources. Given these limits, all known Local Group galaxies except four of the very diffuse, extended dwarf spheroidals located within 100 kpc of the Milky Way would have been detected had they been in our survey. Therefore, we can rule out the possibility that these HVCs are associated with normal but faint dwarf galaxies. If compact HVCs do contain stars, they must have surface brightnesses ≳1 mag arcsec^-2 fainter than most known Local Group galaxies.

Observations of nearby galaxies reveal a strong correlation between the mass of the central dark object M_BH and the velocity dispersion σ of the host galaxy, of the form log(M_BH/M_☉) = α + β log(σ/σ₀); however, published estimates of the slope β span a wide range (3.75-5.3). Merritt & Ferrarese have argued that low slopes (≲4) arise because of neglect of random measurement errors in the dispersions and an incorrect choice for the dispersion of the Milky Way Galaxy. We show that these explanations and several others account for at most a small part of the slope range. Instead, the range of slopes arises mostly because of systematic differences in the velocity dispersions used by different groups for the same galaxies. The origin of these differences remains unclear, but we suggest that one significant component of the difference results from Ferrarese & Merritt's extrapolation of central velocity dispersions to r_e/8 (r_e is the effective radius) using an empirical formula. Another component may arise from dispersion-dependent systematic errors in the measurements. A new determination of the slope using 31 galaxies yields β = 4.02 ± 0.32, α = 8.13 ± 0.06 for σ₀ = 200 km s^-1. The M_BH-σ relation has an intrinsic dispersion in log M_BH that is no larger than 0.25-0.3 dex and may be smaller if observational errors have been underestimated. In an appendix, we present a simple kinematic model for the velocity-dispersion profile of the Galactic bulge.

We present characteristics of the X-ray point-source population in the M84 galaxy observed by the Chandra Advanced CCD Imaging Spectrometer S-Array. We find an excess in the number of sources centered on M84 with a spatial distribution closely corresponding to the M84 stellar light. Given an absence of recent star formation, accreting binaries are the only candidates for the M84 X-ray sources. The majority of M84 sources (with luminosities exceeding 10³⁸ ergs s^-1) exhibit hardness ratios expected from multitemperature blackbody disk emission. The most luminous sources, which we attribute to accreting black holes, exhibit X-ray colors typical of a blackbody spectrum. We also identify the sources whose X-ray colors match the expectations for constituents of the cosmic X-ray background. The number of such sources agrees with that expected to be background sources. After correcting for incompleteness in the source detection, we find a log N- log S for M84 similar to that of the elliptical galaxy NGC 4697, i.e., having a break at a luminosity of L_b = 2.4 × 10³⁸ ergs s^-1, approximately the Eddington limit on the isotropic luminosity for accretion onto a neutron star. The slope of the luminosity function above the break provides evidence for a mass distribution in the M84 accreting black holes.

We analyzed archival Chandra X-ray observations of the central portion of the 30 Doradus region in the Large Magellanic Cloud. The image contains 20 X-ray point sources with luminosities between 5 × 10³² and 2 × 10³⁵ ergs s^-1 (0.2-3.5 keV). A dozen sources have bright WN Wolf-Rayet or spectral type O stars as optical counterparts. Nine of these are within ~3.4 pc of R136, the central star cluster of NGC 2070. We derive an empirical relation between the X-ray luminosity and the parameters for the stellar wind of the optical counterpart. The relation gives good agreement for known colliding-wind binaries in the Milky Way Galaxy and for the identified X-ray sources in NGC 2070. We conclude that probably all identified X-ray sources in NGC 2070 are colliding-wind binaries and that they are not associated with compact objects. This conclusion contradicts earlier studies where it was argued, using ROSAT data, that two earlier discovered X-ray sources are accreting black hole binaries. Five of the 18 brightest stars in R136 are not visible in our X-ray observations. These stars either are single, have low-mass companions, or have very wide orbits. The resulting binary fraction among early-type stars is then unusually high (at least 70%).

We present results from a detailed analysis of a third eclipsing binary (EB) system in the Large Magellanic Cloud, EROS 1044 (~B2 IV-V + ~B2 III-IV). Our study combines the "classical" EB study of light and radial velocity curves with detailed modeling of the observed spectral energy distribution and yields an essentially complete picture of the stellar properties of the system and a determination of its distance. The observational data exploited include optical photometry, space-based UV spectroscopy, and UV/optical spectrophotometry. The advantages of our technique include numerous consistency checks and, in the case of the distance determinations, the absence of zero-point uncertainties and adjustable parameters. We find the EROS 1044 system to consist of a pair of normal, mildly evolved ~21,000 K stars, whose derived properties are consistent with stellar evolution calculations. The distance to the system is 47.5 ± 1.8 kpc. We discuss the implications of our results for three EB systems (HV 2274, HV 982, and EROS 1044) on the general distance to the Large Magellanic Cloud.

We report the first results of a program to study the internal kinematics of globular clusters in the outer halo of the Milky Way. Using the Keck telescope and High Resolution Echelle Spectrometer, we have measured precise radial velocities for 30 candidate red giants in the direction of Palomar 13, an object traditionally cataloged as a compact, low-luminosity globular cluster. We have combined these radial velocities with published proper motion membership probabilities and new CCD photometry from the Keck and Canada-France-Hawaii telescopes to isolate a sample of 21 probable members. We find a systemic velocity of ⟨v_r⟩_s = 24.1 ± 0.5 km s^-1 and a projected, intrinsic velocity dispersion of σ_p = 2.2 ± 0.4 km s^-1. Although modest, this dispersion is nevertheless several times larger than that expected for a globular cluster of this luminosity and central concentration. Taken at face value, it implies a mass-to-light ratio of ϒ_V = 40 based on the best-fit King-Michie model. The surface density profile of Palomar 13 also appears unusual compared to most Galactic globular clusters; depending upon the details of background subtraction and model-fitting, Palomar 13 either contains a substantial population of "extratidal" stars, or is considerably more spatially extended than previously suspected. The full surface density profile is equally well fitted by a King-Michie model having a high concentration and large tidal radius, or by a Navarro-Frenk-White model. We examine—and tentatively reject—a number of possible origins for the observed characteristics of Palomar 13 (e.g., velocity "jitter" among the red giant branch stars, spectroscopic binary stars, nonstandard mass functions, modified Newtonian dynamics) and conclude that the two leading explanations are either catastrophic heating during a recent perigalacticon passage or the presence of a dark matter halo. The available evidence therefore suggests that Palomar 13 is either a globular cluster that is now in the process of dissolving into the Galactic halo or a faint, dark matter-dominated stellar system.

We report the detection of a ringlike H I structure toward l = 900, b = 28, with a velocity of v_LSR = -99 km s^-1. This velocity implies a distance of d = 13 kpc, corresponding to a Galactocentric radius of R_gal = 15 kpc. The l-v_LSR diagram implies an expansion velocity of v_exp ≃ 15 km s^-1 for the shell. The structure has an oblate, irregular shell-like appearance, which surrounds weak infrared emission as seen in the 60 μm Infrared Astronomical Satellite data. At a distance of 13 kpc the size of the object is about 110 × 220 pc and placed 500 pc above the Galactic plane, with a mass of 10⁵M_☉. An expanding shell with such a high mass and diameter cannot be explained by a single supernova explosion or by a single stellar wind bubble. We interpret the structure as a relic of a distant stellar activity region powered by the joint action of strong stellar winds from early-type stars and supernova explosions.

By radiation transfer models we show that the optical properties of grains are poorly constrained by observations of reflection nebulae. The interstellar medium is known to be hierarchically clumped, from a variety of observations (molecules, H I, far-infrared). We have performed radiative transfer through four-tiered, hierarchically clumped dust in a sphere surrounding a central star. Our models have realistic power spectra of the projected density distributions (index ~-3). The input parameters are the albedo (a) and phase parameter (g) of the dust, the radial optical depth of the sphere averaged over all directions (τ₀), and the detailed random distribution of the dust clumps within the sphere. The outputs are the stellar extinction, optical depth, and flux of scattered light as seen from various viewing angles. Observations of a reflection nebula provide the extinction and scattered flux as viewed from one particular direction. Hierarchical geometry has a large effect on the flux of scattered light emerging from a nebula for a particular extinction of the exciting star. There is a very large spread in both scattered fluxes and stellar extinctions for any distribution of dust. Consequently, an observed (τ_ext, τ_sca) can be fitted by a wide range of albedos. There are lower limits on a set by the scattered flux. As an example, in the best-observed reflection nebula, NGC 7023, a(1300 Å) must be higher than ~0.5 if the scattered flux from Witt et al. and a reasonable value for the optical depth within the nebula are adopted. However, the same observations can be fitted with a = 0.8 and 0.6 ≤ g ≤ 0.85, the entire range we considered. With hierarchical geometry it is not completely safe to determine even relative optical constants from multiwavelength observations of the same reflection nebula. The problem is that the geometry effectively changes with wavelength as the opacity of the clumps varies. Limits on the implications of observing the same object in various wavelengths are discussed briefly. Henry uses a recipe to determine the scattered flux from a star with a given extinction. It is claimed to be independent of the geometry. It provides considerably more scattering for given dust optical properties than our models, probably leading to an underestimate of the grain albedos from the UV diffuse Galactic light.

We present 850 μm polarimetry from the James Clerk Maxwell Telescope toward several dense cores within the dark cloud Barnard 1 in Perseus. Significant polarized emission is detected from across the mapped area and is not confined to the locations of bright cores. This indicates the presence of aligned grains and hence a component of the magnetic field in the plane of the sky. Polarization vectors detected away from bright cores are strongly aligned at a position angle of ~90° (east of north), while vectors associated with bright cores show alignments of varying orientations. There is no direct correlation between the polarization angles measured in earlier optical polarimetry toward Perseus and the polarized submillimeter thermal emission. Depolarization toward high intensities is exhibited but toward the brightest core reaches a threshold beyond which no further decrease in polarization percentage is measured. The polarized emission data from the interior envelope are compared with previously published OH Zeeman data to estimate the total field strength and orientation under the assumption of a uniform and nonuniform field component in the region. These results are rough estimates only as a result of the single independent detection of Zeeman splitting toward Barnard 1. The uniform field component is thus calculated to be₀ = 31 μG [±(0.52 - 0.01) - 0.86] in the case in which we have assumed the ratio of the dispersion of the line-of-sight field to the field strength to be 0.2.

The bulk flow velocity for the cluster of interstellar cloudlets within ~30 pc of the Sun is determined from optical and ultraviolet absorption line data, after omitting from the sample stars with circumstellar disks or variable emission lines and the active variable HR 1099. A total of 96 velocity components toward the remaining 60 stars yield a streaming velocity through the local standard of rest of -17.0 ± 4.6 km s^-1, with an upstream direction of l = 23, b = -52 (using Hipparcos values for the solar apex motion). The velocity dispersion of the interstellar matter (ISM) within 30 pc is consistent with that of nearby diffuse clouds, but present statistics are inadequate to distinguish between a Gaussian or exponential distribution about the bulk flow velocity. The upstream direction of the bulk flow vector suggests an origin associated with the Loop I supernova remnant. Groupings of component velocities by region are seen, indicating regional departures from the bulk flow velocity or possibly separate clouds. The absorption components from the cloudlet feeding ISM into the solar system form one of the regional features. The nominal gradient between the velocities of upstream and downstream gas may be an artifact of the Sun's location near the edge of the local cloud complex. The Sun may emerge from the surrounding gas patch within several thousand years.

We examine the role of Alfvén wave damping in heating the plasma in the magnetic funnels of magnetospheric accretion models of young stars. We study four different damping mechanisms of the Alfvén waves: nonlinear, turbulent, viscous-resistive, and collisional. Two different possible origins for the Alfvén waves are discussed: (1) Alfvén waves generated at the surface of the star by the shock produced by the infalling matter and (2) Alfvén waves generated locally in the funnel by the Kelvin-Helmholtz instability. We find that, in general, the damping lengths are smaller than the tube length. Since thermal conduction in the tube is not efficient, Alfvén waves generated only at the star's surface cannot heat the tube to the temperatures necessary to fit the observations. Only for very low frequency Alfvén waves ~10^-5 the ion-cyclotron frequency is the viscous-resistive damping length greater than the tube length. In this case, the Alfvén waves produced at the surface of the star are able to heat the whole tube. Otherwise, local production of Alfvén waves is required to explain the observations. The turbulence level is calculated for different frequencies for optically thin and thick media. We find that turbulent velocites vary greatly for different damping mechanisms, reaching ~100 km s^-1 for the collisional damping of low-frequency waves.

We consider the effects of collective plasma processes on synchrotron emission from highly relativistic electrons. We find, in agreement with the 1970 work of Sazonov, that strong effects are also possible in the absence of a nonrelativistic plasma component, due to the relativistic electrons (and protons) themselves. In contrast with Sazonov, who infers strong effects only in cases in which the ratio of plasma frequency to cyclotron frequency is much larger than the square of the characteristic electron Lorentz factor, ν_p/ν_B ≫ γ, we also find strong effects for 1 ≪ ν_p/ν_B ≪ γ. The modification of the spectrum is prominent at frequencies ν ≤ ν ≡ ν_p min{γ_e, (ν_p/ν_B)^1/2}, where ν generalizes the "Razin-Tsytovich" frequency ν_R ≡ γ_eν_p to the regime ν_p/ν_B ≪ γ. Applying our results to gamma-ray burst (GRB) plasmas, we predict a strong modification of the radio spectrum on a minute timescale following the GRB at the onset of fireball interaction with its surrounding medium, in cases in which the ratio of the energy carried by the relativistic electrons to the energy carried by the magnetic field exceeds ~10⁵. Plausible electron distribution functions may lead to negative synchrotron reabsorption, i.e., to coherent radio emission, which is characterized by a low degree of circular polarization. Detection of these effects would constrain the fraction of energy in the magnetic field, which is currently poorly determined by observations, and moreover would provide a novel handle on the properties of the environment into which the fireball expands.

We present the results of BeppoSAX observations of the young X-ray pulsar PSR J1846-0258, recently discovered at the center of the composite supernova remnant Kes 75. The pulsar (plus nebula) spectrum can be fitted by an absorbed power law with photon index α_ph = 2.16 ± 0.15, N_H = (4.7 ± 0.8) × 10²² cm^-2, and unabsorbed flux ~3.9 × 10^-11 ergs cm^-2 s^-1 (2-10 keV). By joining two observations taken at an interval of 2 weeks, we have been able to obtain a precise measurement of the spin period (P = 324.818968 ± 0.000006 ms). This value, when combined with previous measurements, cannot be fitted by a smooth frequency evolution with a canonical braking index n = 3. With the hypothesis of no glitches and/or significant timing noise, the braking index would be n = 1.86 ± 0.08, and assuming a short initial period, the pulsar age would be ~1700 yr, closer to that of the supernova remnant than the simple estimate τ = P/2 = 723 yr. Other likely possibilities involve the presence of glitches and lead to a wide range of acceptable ages. For example, we obtain n in the range 1.8-2.5 if a glitch occurred near MJD 51500, while for a glitch between 1993 October and 1999 March, we can only get a lower limit of n > 1.89.

We present results from detailed analyses of the ASCA data of the X-ray pulsar GX 301-2 during apastron, periastron, and intermediate orbital phases. We find that the iron Kα emission line possesses a nonzero energy width of 40-80 eV in Gaussian σ during all of the three orbital phases. The observed range in the width is far larger than that expected from velocity shifts in the stellar wind. Quantitative evaluation of the width indicates that the iron Kα line of GX 301-2 is emitted from a region within ~10¹⁰ cm from the neutron star via fluorescence. In addition, we have also discovered a broad wing with σ ≃ 135 eV in the profile of the iron Kα line during the periastron phase. Since the wing profile is modulated according to the spin phase of the pulsar, we conclude that the wing is composed of blue- and redshifted 6.4 keV iron lines emanating from the bipolar accretion columns. We find that, within the accretion column, the He II ionization front is formed at ~10⁷ cm from the pulsar, outside of which iron can be kept in a nearly neutral ionization state. The observed equivalent width of the broad wing component (~100 eV) can be explained by emission in the accretion columns outside the He II ionization front.

We study the dynamics of r-modes in the ocean of a magnetic neutron star. We model the star's ocean with a spherical rotating thin shell and assume that the magnetic field symmetry axis is not aligned to the shell's spin axis. In the magnetohydrodynamic approximation, we calculate the frequency of l = mr-modes in the shell of an incompressible fluid. Different r-modes with l and l ± 2 are coupled by the inclined magnetic field. Kinematical secular effects for the motion of a fluid element in the shell undergoing the l = m = 2 r-mode are studied. The magnetic-corrected drift velocity of a given fluid element undergoing the l = mr-mode oscillations is obtained. The magnetic field increases the magnitude of the fluid drift produced by the r-mode oscillations. The drift velocity is strongly modulated by the inclined magnetic field. We show that the magnetic field is distorted by the high-l magnetic r-modes more strongly than by the low-l modes. Furthermore, because of the shear produced by the r-mode drift velocity, the high-l modes in the ocean fluid will damp faster than the low-l ones.

We investigate the effect of a magnetic field on the global oscillation modes of a rotating fluid star in the magnetohydrodynamic approximation. We present general equations for the modification of any type of fluid mode due to a general magnetic field that is not aligned with the star's spin axis. In the case of any internal dipole magnetic field, we derive the equations for the frequency corrections to the r-modes. We solve for the frequency correction explicitly for the case when the internal dipole field is force-free, including the uniform density case. In the weak-field limit, the spatial form of the r-mode velocity perturbation is unchanged, but the magnitude of the frequency in the rotating frame increases.

The composition of the outer 100 m of a neutron star sets the heat flux that flows outward from the core. For an accreting neutron star in an X-ray transient, the thermal quiescent flux depends sensitively on the amount of hydrogen and helium remaining on the surface after an accretion outburst and on the composition of the underlying ashes of previous H/⁴He burning. Because H/⁴He has a higher thermal conductivity, a larger mass of H/⁴He implies a shallower thermal gradient through the low-density envelope and hence a higher effective temperature for a given core temperature. The mass of residual H and ⁴He varies from outburst to outburst, so the thermal quiescent flux is variable even though the core temperature is constant for timescales ≲10⁴ yr. Heavy elements settle from an H/⁴He envelope in a few hours; we therefore model the quiescent envelope as two distinct layers, H/⁴He over heavier elements, and treat the mass of H/⁴He as a free parameter. We find that the emergent thermal quiescent flux can vary by a factor of 2-3 between different quiescent epochs. The variation is more pronounced at lower interior temperatures, making systems with low quiescent luminosities and frequent outbursts, such as SAX J1808.4-3658, ideal candidates from which to observe this effect. Because the ashes of H/⁴He burning are heavier than ⁵⁶Fe, their thermal conductivity is greatly reduced. This increases the inferred crust temperature beyond previous estimates for a given effective temperature. We survey this effect for different ash compositions and apply our calculations to Cen X-4, Aql X-1, and SAX J1808.4-3658. In the case of Aql X-1, the inferred high interior temperature suggests that neutrino cooling contributes to the neutron star's thermal balance.

We report here the results of a 90 ks BeppoSAX observation of the low-mass X-ray binary and atoll source KS 1731-260 during a quiescent phase. From this observation we derive a source X-ray luminosity of ~10³³ ergs s^-1 (for a source distance of 7 kpc). If the neutron star is spinning at a period of a few milliseconds, as inferred from the nearly coherent oscillations observed during type I X-ray bursts, the quiescent X-ray luminosity constrains the neutron star magnetic field strength. We consider all the mechanisms that have been proposed to explain the quiescent X-ray emission of neutron star X-ray transients and compare the corresponding expectations with the measured upper limit on the X-ray luminosity. We find that, in any case, the neutron star magnetic field is most probably less than ~10⁹ G. We have also observed KS 1731-260, still in its quiescent state, at 1.4 GHz with the Parkes radio telescope to search for radio pulses. We found that no radio signals with millisecond periods are present with an upper limit on the flux of 0.60 mJy using a 4 minute integration time (optimal for a close system with an orbital period smaller than a few hours) and of 0.21 mJy using a 35 minute integration time (optimal for a wide-orbit system).

We obtained Hubble Space Telescope far-ultraviolet spectra of the Z Cam-type dwarf nova RX And during the early rise to outburst and the decline from the same outburst. The spectral wavelength range covered was 1149-1435 Å. The rise spectrum is dominated by strong, very broad absorption lines, while the decline spectrum has strong, narrower absorption with weak to moderately strong emission wings due to the presence of disk material. We have carried out a combined model accretion disk and high-gravity model atmosphere analysis of these spectra. The rise spectrum is best fitted to the Lyα region, metal absorption lines, and longward continuum with a model optically thick accretion disk having an accretion rate of 2 × 10^-10M_☉ yr^-1, a white dwarf of mass M_wd = 0.8 M_☉, with T_eff = 40,000 K, log g = 8.2, and a disk inclination angle of 41°. The hot white dwarf accounts for 62% of the far-UV flux, while the disk accounts for 38%. Our best-fitting multicomponent model of the decline spectrum reveals the presence of a hot (45,000 K) white dwarf contributing 81.4% of the flux during the decline and a remaining accretion disk component with an accretion rate during the decline of 2.85 × 10^-10, contributing 18.6% of the flux. We find that the RX And white dwarf was heated by ~11,000 K during the outburst. Evolutionary simulations with time-variable accretion indicate that boundary layer irradiation has a larger effect on this heating amplitude than compressional heating.

We have observed the prototypical dwarf nova U Gem at quiescence with the Chandra medium- and high-resolution gratings to accomplish the first study of resolved X-ray emission lines in a disk-accreting cataclysmic variable. Doppler tomograms constructed from optical spectra obtained close in time to the X-ray observation show a typical quiescent disk structure with an irradiated secondary but no prominent disk hot spot. The unprecedented spectral resolution of Chandra over past X-ray telescopes reveals prominent narrow emission lines of O, Ne, Mg, Si, S, and Fe. The line fluxes, ratios, and widths indicate that X-ray emission arises from a range of temperatures in a high-density (greater than 10¹⁴ cm^-3) gas, moving at low (less than 300 km s^-1) velocity, with a small (less than 10⁷ cm) scale height compared to the white dwarf radius. Simple models with cooling flows, cooling flows plus isothermal zones, and thermal conduction give reasonable agreement with the low-temperature emission lines, but a good fit to the entire range of lines will require a better understanding of all the parameters that affect the boundary layer.

Our Hubble Space Telescope (HST) ultraviolet spectra of the low mass transfer rate cataclysmic variables EG Cnc and HV Vir at 4 yr and 8 yr past their last superoutburst have shown that the white dwarfs in these systems are among the coolest yet observed in disk-accreting close binaries. The UV-optical fluxes of EG Cnc are consistent with a 11,700-13,000 K white dwarf at a distance of 350-475 pc, while those of HV Vir fit with a 12,500-14,000 K white dwarf at 400-550 pc, with log g = 8.0 ± 0.5 for both stars. In each case, the metal abundance is near 0.3 solar, while the rotation rates are considerably less than breakup. If these systems formed above the period gap and evolved to their short periods in an evolution time of several Gyr, then their white dwarfs are hotter than field white dwarfs of comparable age and hotter than those in close binary systems of similar orbital period that contain magnetic white dwarfs. If these systems have indeed passed the period minimum in their evolution, the implication is that they are significantly heated by the long-term effects of mass transfer from their close companions.

We present the first orbital elements for the massive close binary HD 101131, one of the brightest objects in the young open cluster IC 2944. This system is a double-lined spectroscopic binary in an elliptical orbit with a period of 9.64659 ± 0.00012 days. It is a young system of unevolved stars (approximately 2 million yr old) that are well within their critical Roche surfaces. We use a Doppler tomography algorithm to reconstruct the individual component optical spectra, and we apply well-known criteria to arrive at classifications of O6.5 V((f)) and O8.5 V for the primary and secondary, respectively. We compare the reconstructed spectra of the components to single-star spectrum standards to determine a flux ratio of f₂/f₁ = 0.55 ± 0.08 in the V band. Both components are rotating faster than synchronously. We estimate the temperatures and luminosities of the components from the observed spectral classifications, composite V magnitude, and cluster distance modulus. The lower limits on the masses derived from the orbital elements and the lack of eclipses are 25 and 14 M_☉ for the primary and secondary, respectively. These limits are consistent with the somewhat larger masses estimated from the positions of the stars in the Hertzsprung-Russell diagram and evolutionary tracks for single stars.

We have obtained 11.7 and 17.9 μm images at the Keck I telescope of the circumstellar dust emission from OH 231.8+4.2, an evolved mass-losing red giant with a well-studied bipolar outflow. We detect both a central unresolved point source with a diameter of less than 05 producing F_ν(17.9 μm) = 60 Jy and emission extended more than 1'' away from the star, which is aligned with the bipolar outflow seen on larger scales. We find that the unresolved central source can be explained by an opaque, flared disk with an outer radius of ~5 × 10¹⁵ cm and an outer temperature of ~130 K. One possible model to explain this flaring is that the material in the disk is orbiting the central star and not simply undergoing a radial expansion.

We study the local behavior of gravitational lensing near fold catastrophes. Using a generic form for the lensing map near a fold, we determine the observable properties of the lensed images, focusing on the case in which the individual images are unresolved, i.e., microlensing. Allowing for images not associated with the fold, we derive analytic expressions for the photometric and astrometric behavior near a generic fold caustic. We show how this form reduces to the more familiar linear caustic, which lenses a nearby source into two images that have equal magnification, opposite parity, and are equidistant from the critical curve. In this case, the simplicity and high degree of symmetry allow for the derivation of semianalytic expressions for the photometric and astrometric deviations in the presence of finite sources with arbitrary surface brightness profiles. We use our results to derive some basic properties of astrometric microlensing near folds; in particular, we predict, for finite sources with uniform and limb-darkening profiles, the detailed shape of the astrometric curve as the source crosses a fold. We find that the astrometric effects of limb darkening will be difficult to detect with the currently planned accuracy of the Space Interferometry Mission for Galactic bulge sources; however, this also implies that astrometric measurements of other parameters, such as the size of the source, should not be compromised by an unknown amount of limb darkening. We verify our results by numerically calculating the expected astrometric shift for the photometrically well-covered Galactic binary lensing event OGLE-1999-BUL-23, finding excellent agreement with our analytic expressions. Our results can be applied to any lensing system with fold caustics, including Galactic binary lenses and quasar microlensing.

We show that a space-based gravitational microlensing survey for terrestrial extrasolar planets is feasible in the near future and could provide a nearly complete picture of the properties of planetary systems in our Galaxy. We present simulations of such a survey using a 1-2 m aperture space telescope with a ~2 deg² field of view to continuously monitor ~10⁸ Galactic bulge main-sequence stars. The microlensing techniques allow the discovery of low-mass planets with high signal-to-noise ratio, and the space missions that we have studied are sensitive to planets with masses as low as that of Mars. By targeting main-sequence source stars, which can only be resolved from space, the space-based microlensing survey is able to detect enough light from the lens stars to determine the spectral type of one-third of the lens stars with detected planets, including virtually all of the F, G, and K stars, which comprise one-quarter of the event sample. This enables the determination of the planetary masses and separations in physical units as well as the abundance of planets as a function of stellar type and distance from the Galactic center. We show that a space-based microlensing planet search program has its highest sensitivity to planets at orbital separations of 0.7-10 AU, but it will also have significant sensitivity at larger separations and will be able to detect free-floating planets in significant numbers. This complements the planned terrestrial planet transit missions, which are sensitive to terrestrial planets at separations of ≤1 AU. Such a mission should also detect ~50,000 giant planets via transits, and it is, therefore, the only proposed planet detection method that is sensitive to planets at all orbital radii.

The solar system gas giant planets are oblate due to their rapid rotation. A measurement of the planet's projected oblateness would constrain the planet's rotational period. Planets that are synchronously rotating with their orbital revolution will be rotating too slowly to be significantly oblate; these include planets with orbital semimajor axes ≲0.2 AU (for M_P ~ M_Jupiter and M_* ~ M_☉). Jupiter-like planets in the range of orbital semimajor axis 0.1 to 0.2 AU will tidally evolve to synchronous rotation on a timescale similar to main-sequence stars' lifetimes. In this case, an oblateness detection will help constrain the planet's tidal Q value.

The projected oblateness of a transiting extrasolar giant planet is measurable from a very high photometric precision transit light curve. For a Sun-sized star and a Jupiter-sized planet, the normalized flux difference in the transit ingress/egress light curve between a spherical and an oblate planet is a few to 15 × 10^-5 for oblateness similar to Jupiter and Saturn, respectively. The transit ingress and egress are asymmetric for an oblate planet with orbital inclination different from 90° and a nonzero projected obliquity. A photometric precision of 10^-4 has been reached by Hubble Space Telescope (HST) observations of the known transiting extrasolar planet HD 209458b.

Kepler, a NASA discovery-class mission designed to detect transiting Earth-sized planets, requires a photometric precision of 10^-5 and expects to find 30 to 40 transiting giant planets with orbital semimajor axes <1 AU, about 20 of which will be at >0.2 AU. Furthermore, part-per-million photometric precision (reached after averaging over several orbital periods) is expected from three other space telescopes to be launched within the next three years. Thus, an oblateness measurement of a transiting giant planet is realistic in the near future.

This paper discusses a quasi-static evolution of a force-free magnetic field under slow sheared footpoint motions on the plasma's boundary, an important problem with applications to the solar and accretion disk coronae. The main qualitative features of the evolution (such as field-line expansion and opening) are considered, and a comparison is made between two different geometrical settings: the Cartesian case with translational symmetry along a straight line, and the axisymmetric case with axial symmetry around the rotation axis. The main question addressed in the paper is whether a continuous sequence of force-free equilibria describes the evolution at arbitrarily large values of the footpoint displacement, or the sequence ends abruptly and the system exhibits a loss of equilibrium at a finite footpoint displacement. After a formal description of the problem, a review/discussion of the extensive previous work on the subject is given. After that, a series of simple scaling-type arguments, explaining the key essential reason for the main qualitative difference between the two geometry types, is presented. It is found that, in the Cartesian case, force-free equilibria exist at arbitrarily large values of shear and the field approaches the open state only at infinite shear, whereas in the axisymmetric case the field opens up already at a finite shear.

Solar coronal sigmoidal active regions have been shown to be precursors to some coronal mass ejections. Sigmoids, or S-shaped structures, may be indicators of twisted or helical magnetic structures, having an increased likelihood of eruption. We present here an analysis of a sigmoidal region's three-dimensional structure and how it evolves in relation to its eruptive dynamics. We use data taken during a recent study of a sigmoidal active region passing across the solar disk (an element of the third Whole Sun Month campaign). While S-shaped structures are generally observed in soft X-ray (SXR) emission, the observations that we present demonstrate their visibility at a range of wavelengths including those showing an associated sigmoidal filament. We examine the relationship between the S-shaped structures seen in SXR and those seen in cooler lines in order to probe the sigmoidal region's three-dimensional density and temperature structure. We also consider magnetic field observations and extrapolations in relation to these coronal structures. We present an interpretation of the disk passage of the sigmoidal region, in terms of a twisted magnetic flux rope that emerges into and equilibrates with overlying coronal magnetic field structures, which explains many of the key observed aspects of the region's structure and evolution. In particular, the evolving flux rope interpretation provides insight into why and how the region moves between active and quiescent phases, how the region's sigmoidicity is maintained during its evolution, and under what circumstances sigmoidal structures are apparent at a range of wavelengths.

Using advanced instrumentation on the ACE spacecraft, we have conducted a survey of solar energetic particle spectra in ³He-rich events over a broad energy range ~80 keV nucleon^-1 to 15 MeV nucleon^-1 during the period 1997 September-2001 March. The spectra of ⁴He and heavy ions (C, N, O, Ne, Mg, Si, S, Ca, Fe) were generally similar over this range but often hardened below ~1 MeV nucleon^-1. In most of the events there was even stronger hardening of the ³He spectrum below ~1 MeV nucleon^-1, leading to an energy-dependent ³He : ⁴He ratio. These observations point to unique and distinct properties of ³He in these events and place new constraints on models that seek to explain enhancements of ³He and heavy ions using the same mechanisms. In addition to the events with spectra in the form of power laws or double power laws, there is a second class of event in which the low-energy ³He and Fe spectra are rounded, while the ⁴He remains a power law. In these cases ³He and Fe spectra can be fitted at low energies by a stochastic acceleration model, but this model does not explain the higher energy portions of these spectra, nor the power-law spectral forms of the ⁴He. These observations appear to require an additional mechanism, such as acceleration by cascading MHD turbulence. The ³He enrichment pattern that we observe suggests that all these different spectral features might be due to processes with a common origin but then followed by different acceleration histories.

Solar flare emission was measured at 212 GHz in the submillimeter range by the Submillimeter Solar Telescope in the 1.2-18 GHz microwave range by the Owens Valley Solar Array and in the gamma-ray energy range (continuum) by experiments on board the Yohkoh (>1.2 MeV) and Shenzhou 2 (>0.2 MeV) satellites. At the burst onset, the submillimeter and microwave time profiles were well correlated with gamma rays to the limit of the temporal resolution (≤10 s). At 212 GHz, fast pulses (<1 s), defined as time structures in excess of the bulk emission, were identified as the flux increased. Their spatial positions were scattered by tens of arcseconds with respect to the main burst emission position. Correlation of submillimeter emission with gamma-ray fast time structures shorter than 500 ms is suggested at the gamma-ray maximum. The time variation of the rate of occurrence of the submillimeter rapid pulses was remarkably well correlated with gamma-ray intensities in the energy range (>1.2 MeV), attaining nearly 50 pulses per minute at the maximum. These results suggest that gamma rays might be the response to multiple rapid pulses at 212 GHz and might be produced at different sites within the flaring region.

We present observational evidence that the occurrence of homologous flares in an active region is physically related to the injection of magnetic helicity by horizontal photospheric motions. We have analyzed a set of 1 minute cadence magnetograms of NOAA AR 8100 taken over a period of 6.5 hr by the Michelson Doppler Imager on board the Solar and Heliospheric Observatory. During this observing time span, seven homologous flares took place in the active region. We have computed the magnetic helicity injection rate into the solar atmosphere by photospheric shearing motions and found that a significant amount of magnetic helicity was injected during the observing period. In a strong M4.1 flare, the magnetic helicity injection rate impulsively increased and peaked at the same time as the X-ray flux. The flare X-ray flux integrated over the X-ray emission time strongly correlates with the magnetic helicity injected during the flaring interval. The integrated X-ray flux is found to be a logarithmically increasing function of the injected magnetic helicity. Our results suggest that injection of helicity and abrupt increase of helicity magnitude play a significant role in flare triggering.

By combining Yohkoh soft X-ray images with high-resolution magnetograms simultaneously obtained at La Palma, we studied photospheric magnetic signatures responsible for soft X-ray microflares (active-region transient brightenings). In order to have a reliable correspondence between the photosphere and the corona, we studied 16 pointlike transient brightenings with X-ray source size less than 10'' occurring during periods when the seeing was excellent at La Palma, although a lot of transient brightenings were in forms of multiple- or single-loop structures. In half of the studied events, small-scale emergences of magnetic flux loops are found in the vicinity of the transient brightenings. Six events of that half show that a small-scale flux emergence accompanies the X-ray brightening 5-30 minutes prior to its onset. In the other half of the studied events, no apparent evolutionary change of magnetic flux elements is found associated with the transient brightenings. Many of these events are found in rather strong magnetic fields, such as sunspots and pores, implying that small-scale changes of magnetic flux are obscured or suppressed by strong magnetic fields. The horizontal plasma flows derived from local cross-correlation tracking of granules in continuum images are suppressed at the feet of some X-ray transient brightenings.

We have observed the narrow-line regions (NLRs) of the seven brightest radio-quiet Palomar-Green (or Bright Quasar Survey) quasars (z < 0.5) with the Wide Field Planetary Camera 2 on board the Hubble Space Telescope (HST). Linear-ramp filters were used to image the [O III] λ5007 line emission with 00455-01 pixel resolution. We find that the NLRs are very compact with typical extents of 2''-4''. Two quasars show compact filamentary structures similar to Seyfert NLRs. They may be related to radio outflows. Most interestingly, when including a sample of Seyfert galaxies observed with HST, we tentatively find that the size of the NLR is proportional to the square root of the [O III] luminosity. This is comparable to the scaling found for the size of the broad-line region with continuum luminosity, which has been interpreted in terms of a constant photoionization parameter. The relation determined here connects the NLR of radio-quiet quasars and Seyfert galaxies over 3 orders of magnitude in [O III] luminosity.

The expansion of the universe leads to a rapid drop in the hydrogen Lyα effective optical depth of the intergalactic medium (IGM), _eff ∝ (1 + z)^3.8, between redshifts 4 and 3. Measurements of the temperature evolution of the IGM and of the He II opacity both suggest that He II reionizes in this redshift range. We use hydrodynamical simulations to show that the observed temperature increase associated with He II reionization leads to a relatively sudden decrease in _eff around the reionization epoch of ≈10%. We find clear evidence for such a feature in the evolution of _eff determined from a sample of ~1100 quasars obtained from the Sloan Digital Sky Survey. He II reionization starts at redshift ≈3.4 and lasts for Δz ≈ 0.4. The increase in the IGM temperature also explains the widths of hydrogen absorption lines as measured in high-resolution spectra.

We have conducted pointed redshift surveys for galaxies in the direction of bright active galactic nuclei whose Hubble Space Telescope far-UV spectra contain nearby (cz ≲ 30,000 km s^-1), low column density [12.5 ≤ log N(cm^-2) ≤ 14.5] Lyα forest absorption systems. Here we present results for four lines of sight that contain nearby (cz ≲ 3000 km s^-1) Lyα absorbers in galaxy voids. Although our data go quite deep [-13 ≤ M_B(limit) ≤ -14] out to impact parameters of ρ ≲ 100-250 h kpc, these absorbers remain isolated and thus appear to be truly intergalactic, rather than part of galaxies or their halos. Since we and others have discovered no galaxies in voids, the only baryons detected in the voids are in the Lyα "clouds." Using a photoionization model for these clouds, the total baryonic content of the voids is 4.5% ± 1.5% of the mean baryon density.

We have measured the Sunyaev-Zeldovich (SZ) effect toward the Coma Cluster (A1656) with the Millimetre and Infrared Testagrigia Observatory experiment, a 2.6 m telescope equipped with a four-channel 17' (FWHM) photometer. Measurements at frequency bands 143 ± 15, 214 ± 15, 272 ± 16, and 353 ± 13 GHz were made during 120 drift scans of Coma. We describe the observations and data analysis that involved extraction of the SZ signal by employing a spatial and spectral decorrelation scheme to remove a dominant atmospheric component. The deduced values of the thermal SZ effect in the first three bands are ΔT₀ = -179 ± 38, -33 ± 81, and 170 ± 35 μK in the cluster center. The corresponding optical depth, τ = (4.1 ± 0.9) × 10^-3, is consistent (within errors) with both the value from a previous low-frequency SZ measurement and the value predicted from the X-ray-deduced gas parameters.

We present results from a simultaneous Chandra High Energy Transmission Grating and XMM-Newton observation of NGC 3516. We find evidence of several narrow components of Fe Kα along with a broad line. We consider the possibility that the lines arise in a blob of material ejected from the nucleus with velocity ~0.25c. We also consider an origin in a neutral accretion disk, suffering enhanced illumination at 35R_g and 175R_g, perhaps as a result of magnetic reconnection. The presence of these narrow features indicates there is no Comptonizing region along the line of sight to the nucleus. This in turn is compelling support for the hypothesis that broad Fe Kα components are, in general, produced by strong gravity.

We present new spectroscopic observations of the gravitational arcs and the brightest cluster galaxy (BCG) in the cluster MS 2137-23 (z = 0.313) obtained with the Echelle Spectrograph and Imager on the Keck II telescope. We find that the tangential and radial arcs arise from sources at almost identical redshifts (z = 1.501, 1.502). We combine the measured stellar velocity dispersion profile of the BCG with a lensing analysis to constrain the distribution of dark and stellar matter in the central 100 kpc of the cluster. Our data indicate a remarkably flat inner slope for the dark matter profile, ρ_d ∝ r^-β, with β < 0.9 at a 99% CL. Steep inner slopes obtained in cold dark matter cosmological simulations—such as Navarro, Frenk, & White (β = 1) or Moore (1.5) universal dark matter profiles—are ruled out at better than 99% CL. As baryon collapse is likely to have steepened the dark matter profile from its original form, our data provide a powerful test of the cold dark matter paradigm at the cluster mass scale.

We present an initial analysis of a new XMM-Newton observation of NGC 1399, the central elliptical galaxy of the Fornax group. Spectral fitting of the spatially resolved spectral data of the European Photon Imaging Camera MOS and pn CCDs reveals that a two-temperature (2T) model of the hot gas is favored over single-phase and cooling flow models within the central ~20 kpc. The preference for the 2T model applies whether or not the data are deprojected. The cooler component has a temperature (~0.9 keV) similar to the kinetic temperature of the stars, while the hotter component has a temperature (~1.5 keV) characteristic of the virial temperature of a ~10¹³M_☉ halo. The two-phase model (and other multitemperature models) removes the "Fe bias" within r ≲ 20 kpc and gives Z_Fe/Z_☉ ≈ 1.5-2. At larger radii, the iron abundance decreases until Z_Fe/Z_☉ ~ 0.5 for r ~ 50 kpc. The Si abundance is supersolar (1.2-1.7 solar) within the central regions, while Z_Si/Z_Fe ≈ 0.8 over the entire region studied. These Fe and Si abundances imply that ≈80% of the Fe mass within r ~ 50 kpc originates from Type Ia supernovae (SNe Ia). This SNe Ia fraction is similar to that inferred for the Sun and therefore suggests a stellar initial mass function similar to the Milky Way.

Type I X-ray bursts are very common in neutron star X-ray binaries, but no type I burst has been seen in the dozen or so binaries in which the accreting compact star is too massive to be a neutron star and is therefore identified as a black hole candidate. We have carried out a global linear stability analysis of the accumulating fuel on the surface of a compact star to identify the conditions under which thermonuclear bursts are triggered. Our analysis, which improves on previous calculations, reproduces the gross observational trends of bursts in neutron star systems. It further shows that, if black hole candidates have surfaces, they would very likely exhibit instabilities similar to those that lead to type I bursts on neutron stars. The lack of bursts in black hole candidates is thus significant and indicates that these objects have event horizons. We discuss possible caveats to this conclusion.

We discuss the nature of the infrared (IR) counterpart of GX 17+2, one of the most luminous of the persistently bright X-ray binaries. Chandra HRC-S astrometry is consistent with either NP Ser (the original counterpart of GX 17+2 proposed by M. Tarenghi & C. Reina in 1972) or star "A" of E. W. Deutsch et al. as the counterpart of the X-ray source. However, we present Keck K-band observations that reveal a bright counterpart in the radio error circle of Deutsch et al. 09 north of NP Ser itself. Furthermore, the position of this counterpart is consistent with that of star A to within 01, implying an amplitude of variation of ~25-33 between the Keck observations and the Hubble Space Telescope (HST) measurements of Deutsch et al. Subsequent Keck imaging also reveals star A in an "IR-faint" state (K = 18.3 mag, with a corresponding amplitude of variability of ~22). In addition, archival Cerro Tololo Inter-American Observatory observations provide evidence for K-band variability, albeit of smaller amplitude. The HST and Keck K-band variations, however, do not appear to be accompanied by any changes in the overall X-ray luminosity of GX 17+2 as measured by contemporaneous (but not simultaneous) Rossi X-Ray Timing Explorer All-Sky Monitor observations. We propose instead that the large radio outbursts observed when the source is in the horizontal branch of its "Z" state are likely to give rise to synchrotron flares in the IR. The amplitude of the radio flares is in agreement with this scenario. Such IR variability, unrelated (directly) to X-ray reprocessing and the gross characteristics of the mass accretion rate, may be present in the IR flux of other low-mass X-ray binaries but harder to see owing to the intrinsically brighter IR fluxes of the longer period systems.

Using the StarTrack population synthesis code, we analyze the possible formation channels of the Galactic microquasar GRS 1915+105, harboring the most massive known stellar black hole. We find that the evolution and the current state of GRS 1915+105 can be reproduced in a model with lowered wind mass-loss rates in massive stars. We conclude that the existence of GRS 1915+105 implies that stellar winds in massive stars are a factor of 2 weaker—but still within observational bounds—than those used in stellar models. We discuss the mass transfer initiated by the massive primary in binaries, which are the possible precursors of systems harboring massive black holes, and show that it is not clearly either case B or C but is instead initiated during the core helium-burning phase of the massive donor star, the progenitor of the black hole. We also argue that massive black holes do not receive any kicks at their formation and are formed through direct collapse of the progenitor star without an accompanying supernova explosion.

The deepest optical image ever in a globular star cluster, a Hubble Space Telescope 123 orbit exposure in a single field of Messier 4, was obtained in two filters (F606W, F814W) over a 10 week period in early 2001. A somewhat shallower image obtained in 1995 allowed us to select out cluster and field objects via their proper-motion displacement, resulting in remarkably clean color-magnitude diagrams that reach to V = 30, I = 28. The cluster main-sequence luminosity function contains very few stars fainter than M_V = 15.0, M_I = 11.8, which, in both filters, is more than 2 mag brighter than our limit. This is about the faintest luminosity seen among field Population II subdwarfs of the same metallicity. However, there remains a sprinkling of potential cluster stars to lower luminosity all the way down to our limiting magnitudes. These latter objects are significantly redder than any known metal-poor field subdwarf. Comparison with the current generation of theoretical stellar models implies that the masses of the lowest luminosity cluster stars observed are near 0.09 M_☉. We derive the mass function of the cluster in our field and find that it is very slowly rising toward the lowest masses with no convincing evidence of a turnover even below 0.1 M_☉. The formal slope between 0.65 and 0.09 M_☉ is α = 0.75 (Salpeter of 2.35) with a 99% confidence interval of 0.55-1.05. A consistency check between these slopes and the number of observed cluster white dwarfs yields a range of possible conclusions, one of which is that we have indeed seen the termination of the white dwarf cooling sequence in M4.

We present the white dwarf sequence of the globular cluster M4, based on a 123 orbit Hubble Space Telescope exposure, with a limiting magnitude of V ~ 30 and I ~ 28. The white dwarf luminosity function rises sharply for I > 25.5, consistent with the behavior expected for a burst population. The white dwarfs of M4 extend to approximately 2.5 mag fainter than the peak of the local Galactic disk white dwarf luminosity function. This demonstrates a clear and significant age difference between the Galactic disk and the halo globular cluster M4. Using the same standard white dwarf models to fit each luminosity function yields ages of 7.3 ± 1.5 Gyr for the disk and 12.7 ± 0.7 Gyr for M4 (2 σ statistical errors).

Isolated low-mass stars are formed, in the standard picture, from the collapse of dense cores condensed out of strongly magnetized molecular clouds. The dynamically collapsing inflow traps nearly half of the critical magnetic flux needed for the core support and deposits it in a small region surrounding the protostar. It has been argued previously that the deposited flux can slow down the inflow, allowing matter to pile up and settle along field lines into a magnetically supported, circumstellar disk. Here we show that the disk typically contains ~10% of the stellar mass and that it could become self-gravitating under plausible conditions during the rapidly accreting, "Class 0" phase of star formation. Subsequent fragmentation of the self-gravitating, magnetically subcritical disk, driven by magnetic diffusion, could produce fragments of substellar masses, which collapse to form brown dwarfs and possibly massive planets. This scenario predicts substellar object formation at distances of order 100 AU from the central star, although orbital evolution is possible after formation. It may provide an explanation for the small, but growing, number of brown dwarf companions found around nearby stars by direct imaging. The relatively large formation distances make the substellar companions vulnerable to dynamic ejection, particularly in binary (multiple) systems and dense clusters. Those ejected may account for, at least in part, the isolated brown dwarfs and perhaps free-floating planetary mass objects.

We present images of the J = 10-9 rotational lines of HC₃N in the vibrationally excited levels 1v7, 1v6, and 1v5 of the hot core (HC) in Orion KL. The images show that the spatial distribution and the size emission from the 1v7 and 1v5 levels are different. While the J = 10-9 1v7 line has a size of 4'' × 6'' and peaks 11 northeast of the 3 mm continuum peak, the J = 10-9 1v5 line emission is unresolved (<3'') and peaks 13 south of the 3 mm peak. This is a clear indication that the HC is composed of condensations with very different temperatures (170 K for the 1v7 peak and >230 K for the 1v5 peak). The temperature derived from the 1v7 and 1v5 lines increases with the projected distance to the suspected main heating source I. Projection effects along the line of sight could explain the temperature gradient as being produced by source I. However, the large luminosity required for source I (>5 × 10⁵L_☉) to explain the 1v5 line suggests that external heating by this source may not dominate the heating of the HC. Simple model calculations of the vibrationally excited emission indicate that the HC can be internally heated by a source with a luminosity of 10⁵L_☉, located 12 southwest of the 1v5 line peak (18 south of source I). We also report the first detection of high-velocity gas from vibrationally excited HC₃N emission. Based on excitation arguments, we conclude that the main heating source is also driving the molecular outflow. We speculate that all the data presented in this Letter and the IR images are consistent with a young massive protostar embedded in an edge-on disk.

We show that the physical conditions in CRL 618 are such that efficient formation of benzene, C₆H₆, occurs. A combination of high temperatures, high densities, and high ionization rates drives an efficient ion-molecule chemistry involving condensation reactions of acetylene and its derivatives, rather than reactions involving atomic hydrogen, as was suggested for the interstellar synthesis of benzene. We find a column density of benzene within a factor of 2 of that observed providing that the material is trapped in a long-lived reservoir of gas in the disk around CRL 618. We note that the chemistry can give rise to other carbon chain molecules as well as a large abundance of benzonitrile, C₆H₅CN.

We report the detection of fully resolved absorption lines of A-X bands from interstellar ¹²C¹⁷O and ¹²C¹⁸O, through high-resolution spectroscopy of X Persei with the Space Telescope Imaging SpectrographBased on observations obtained with the NASA/ESA Hubble Space Telescope through the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS 5-26555.. The first ultraviolet measurement of an interstellar ¹²C¹⁷O column density shows that its isotopomeric ratio is ¹²C¹⁶O/¹²C¹⁷O = 8700 ± 3600. Simultaneously, the second ultraviolet detection of interstellar ¹²C¹⁸O establishes its isotopomeric ratio at 3000 ± 600. These ratios are about five times higher than local ambient oxygen isotopic ratios in the ISM. Such severe fractionation of rare species shows that both ¹²C¹⁷O and ¹²C¹⁸O are destroyed by photodissociation, whereas ¹²C¹⁶O avoids destruction through self-shielding. This is to be contrasted with our ratio of ¹²C¹⁶O/¹³C¹⁶O = 73 ± 12 toward X Per, which is indistinguishable from ¹²C/¹³C, the result of a balance between the photodissociation of ¹³C¹⁶O and its preferential formation via the isotope exchange reaction between CO and C⁺.

We review the behavior of the oscillating shear layer produced by gravity waves below the surface convection zone of the Sun. We show that, under asymmetric filtering produced by this layer, gravity waves of low spherical order that are stochastically excited at the base of the convection zone of late-type stars can extract angular momentum from their radiative interior. The timescale for this momentum extraction in a Sun-like star is on the order of 10⁷ yr. The process is particularly efficient in the central region, and it could produce there a slowly rotating core.

In typical observations of coronal mass ejections (CMEs), a magnetic structure of a helmet-shaped closed configuration bulges out and eventually opens up. However, a spontaneous transition between these field configurations has been regarded to be energetically impossible in force-free fields according to the Aly-Sturrock theorem. The theorem states that the maximum energy state of force-free fields with a given boundary normal field distribution is the open field. The theorem implicitly assumes the existence of the maximum energy state, which may not be taken for granted. In this study, we have constructed force-free fields containing tangential discontinuities in multiple flux systems. These force-free fields can be generated from a potential field by footpoint motions that do not conserve the boundary normal field distribution. Some of these force-free fields are found to have more magnetic energy than the corresponding open fields. The constructed force-free configurations are compared with observational features of CME-bearing active regions. Possible mechanisms of CMEs are also discussed.

We report the results of the low-dispersion spectroscopic observations of comet Ikeya-Zhang (C/2002 C1) performed from 2002 March 10 to 20. The unidentified molecular bands that have been recognized in the plasma tail of several comets are detected in an antisunward coma of the comet Ikeya-Zhang. Our observations show the flux of unidentified bands at 5310 Å is correlated to the flux of H₂O⁺ as reported for three comets by S. Wyckoff et al. The observed column density ratio between H₂O⁺ and CO⁺, and the flux ratio between the unidentified bands and CO⁺ varied day by day, by a factor of ≈2 in our observations. However, it appears that the ratios are proportional to each other. We conclude that a parent of unidentified bands is produced from or generates H₂O⁺ directly or indirectly. We propose the hypothesis that H₂O⁺ is the parent of the unidentified bands since similar structures of emission bands are recognized in some laboratory studies on charge transfer collisions between neutral water and Ar⁺ or N.