Lucia Munoz-Franco

We present a comprehensive study of the properties of nearly coherent brightness oscillations in a large sample of type-I X-ray bursts observed from low mass X-ray binary 4U 1728-34. We have analyzed ˜545 ks of data of this source obtained with the Rossi X-ray Timing Explorer over a 3 year span from 1996--1999. The data contain 38 bursts, 16 of which show oscillations. We find no burst oscillations present when the inferred mass accretion rate of the system is lowest. Furthermore, we define a measure of the strength of the oscillations and find that this strength increases with increasing mass accretion rate. This correlation is particularly evident within bursts detected only a few weeks apart, and becomes less clear for bursts separated in time by several months to years. The correlation we find for 4U 1728-34 between the burst oscillations and the inferred accretion rate of the system is similar to that found for KS 1731-260 by Muno et al., where the burst oscillations are only present at relatively high mass accretion rates. However, unlike the case for KS 1731-260, we find an anticorrelation in the bursts of 4U 1728-34 between the existence of episodes of photospheric radius expansion and the inferred mass accretion rate of the system. Moreover, we distinguish between burst oscillations present in the rise and in the decay phase of the burst and find that the bursts with oscillations only in the decay phase occur at intermediate accretion rates, while those with oscillations in both the rise and the decay phases are concentrated at high inferred mass accretion rates. We discuss these results in the context of the theory of thermonuclear bursts and propose intrinsic differences in the neutron stars in KS 1731-260 and 4U 1728-34 as the origin of the different behaviors.

Yuan-Nan Young

Slot convection refers to buoyantly driven flow due to the lateral temperature difference. In double-diffusive slot convection the destabilizing temperature gradient (transverse to the gravitational direction) is competing with a stabilizing concentration gradient. For suitable combination of physical parameters, layers form as a result of buoyancy balance between the lateral thermal difference and the stabilizing concentration gradient. In chapter 2, we directly simulate this system using a two- dimensional pseudospectral code. Incompressibility is achieved by the consistent implementation of the tau- correction. We find that layer dynamics depends on the particulars of the imposed boundary conditions for the temperature at the sidewalls and the density stratification ratio (the relative strength of the stabilizing solute gradient to the destabilizing horizontal thermal difference). We demonstrate the effects of the density stratification ratio on the layer dynamics for the constant sidewall temperature case, and we also study the case of constant lateral heat flux in order to understand the effects of the temperature boundary conditions. We apply the argument for layering in turbulent stratified fluids to our problem, and find-despite the tilted nature of cell boundaries in our case-similarities in both the averaged equations and actual layer evolution. Finally, we provide details for both edge mergers and interior mergers. In chapter 3, we investigate the miscible Rayleigh-Taylor (RT) instability in both 2 and 3 dimensions using direct numerical simulations, where the working fluid is assumed incompressible under the Boussinesq approximation. With a variety of diagnostics, we develop a physical picture for the detailed temporal development of the mixed layer: We identify three distinct evolutionary phases in the development of the mixed layer. Our analysis provides an explanation for the observed differences between two and three-dimensional RT instability; the analysis also leads us to concentrate on the RT models which (1)work equally well for both laminar and turbulent flows, and (2)do not depend on turbulent scaling within the mixing layer between fluids. These candidate RT models are based on point sources within bubbles (or plumes) and interaction with each other (or the background flow). With this motivation, we examine the evolution of single plumes, and relate our numerical results (of single plumes) to a simple analytical model for plume evolution.

Michael Joffre

We present weak lensing mass map reconstructions of three nearby ( zcd < 0.06) clusters in various stages of merging: Abell 3266, Abell 3667, and Abell 3158. The weak lensing shear in all clusters was well fit by a Navarro, Frenk, and White profile [Navarro et al. (1997)]. Additionally, we used the lensing shear to predict the rc and beta values of a beta model fit to the X-ray intensity under assumptions of isothermality and hydrostatic equilibrium. The most relaxed cluster, Abell 3158, reproduces these values found in X-ray analysis of the cluster, while the merging clusters' parameters are farther from those found through X-ray analysis. We find total cluster masses in agreement with vith X-ray estimates and slightly lower than virial mass estimates. Substructure in the reconstruction is also measured quantitatively in all clusters, showing strong differences between the merging clusters and the relaxed cluster. These clusters form the first three results from an X-ray luminosity-selected weak lensing survey of 19 low-redshift (zcl < 0.1) southern clusters.

Gordon Richards

Moderate resolution spectra were recorded for 24 radio selected quasars (z ~ 2.5) discovered in the FIRST VLA survey. This work is motivated by the study of a heterogeneous set of QSO absorption line spectra which showed that as many as 36% of the absorbers normally thought to be intergalactic are correlated with physical properties of the background QSO (Richards et al. 1999). Spectra were taken for these 24 quasars in order to test this finding on a more homogeneous data set. The spectra have been searched for the most frequently observed absorption lines. The absorption line properties are compared to the radio properties of the quasars. The primary results are summarized as follows: (1)The excess of C IV absorbers in flat-spectrum quasars as compared to steep-spectrum quasars is confirmed. This excess does not seem to be the result of a bias in the quality of the QSO spectra or the QSO redshift. (2)A less significant excess of high- velocity C IV absorbers is seen in radio- quiet QSOs as compared to radio-loud QSOs. (3)The Sloan Digital Sky Survey QSO spectra will allow for the detection of a high-velocity, narrow population of intrinsic absorbers as small as 7%. (4)Though C IV is more prevalent than Mg II, high-velocity C IV absorbers in steep-spectrum quasars are observed only as often as Mg II. The true frequency of C IV absorption by intervening galaxies may be closer to that of Mg II. (5)The ``associated'' absorption population has a HWHM of ~1200kms-1 as measured from the tail of ``infalling'' absorbers. These infalling absorbers cannot be reconciled by offsets from the systemic redshift. (6)The associated systems display a tail of ``ejected'' material at blueshifts up to 6000 km s-1. This feature is not easily accounted for by a virialized population and argues for a significant contribution by ejected material. (7)The velocity distribution of associated absorbers is well centered at zero. This finding is at odds with the evidence that C IV emission redshifts are typically too small. This offset from systemic may result from C IV absorbers being tied to the C IV BELR.

Jeffrey Jewell

While the power spectrum of the cosmic microwave background (CMB) is parameterized by the underlying cosmological model, statistical characterizations of the CMB beyond the power spectrum probe the physical mechanism for generating the initial density perturbations, provide details about secondary processes along the line of sight, and provide a systematic check of residual, non-Gaussian foreground emission. An information theoretic approach to the problem of characterizing non-Gaussian random fields is developed, and then applied to a statistical characterization of the galactic dust as a non-Gaussian foreground for the CMB. The Kullback-Leibler (KL) distance between histograms of wavelet coefficients for selected IRAS patches and corresponding random phase realizations is found to give a highly significant detection of the non-Gaussianity. The KL distance is also compared to the skewness and kurtosis of the IRAS and random phase wavelet coefficients, and the efficiency of these statistics quantified by the rate at which the detection of non- Gaussianity is lost as a function of the fraction of phases randomized.

Luisa Rebull

We have photometrically monitored ~3600 young, low- mass stars in four 45' × 45' fields in the outer Orion Nebula Cluster (ONC), surrounding but not including the Trapezium region. The ~300 periodic variables we find do not produce the expected bimodal distribution. There is no unambiguous correlation of period with ( IC - Ks), ( H - Ks), and (U - V) color excesses or more indirect disk indicators; the slowest rotators are not necessarily the disk candidates, and the disk candidates are not necessarily the slow rotators, regardless of how one defines a disk candidate. To the extent that the small numbers allow, the disk candidates represent a constant fraction of the total sample to P = 15d, beyond which there are no disk candidates, inconsistent with the hypothesis that the more slowly rotating stars are more likely to have disks. We find an intriguing relationship between specific angular momentum (j) and some excesses, namely ( H - Ks) and the strongest ( U - V); whereas stars are found with j values ranging over ~106 - 108 km2s-1, disk candidates with these excesses are restricted to 106.5 < j < 107.5 km2s-1. A similar relationship is not found for the (IC - Ks) disk candidates, nor in the less excessive ( U - V) candidates. There is no clear correlation between signal amplitude and period or log j. Other investigators have found differences in period distributions for stars more and less massive than 0.25 Msolar ; we find ambiguous evidence for differences in distributions of P and find no difference in those of log j on either side of this boundary, although for stars more massive than ~0.6 Msolar , values appear to cluster around j = 107 km2s-1. When comparing the log j distribution derived here to those for other clusters, we find that it is consistent with a population of stars draining angular momentum into disks. We conclude that disk locking may be operating, but it is not the complete solution to the problem of angular momentum distributions in young stars.

Scott Severson

The South Pole Infrared Explorer (SPIREX) is the name given a series of experiments to study and exploit the low thermal background at the South Pole for near- infrared observations. We observed the collision of the fragments of Comet Shoemaker-Levy 9 with Jupiter using the SPIREX 60 cm bent Cassegrain telescope. The camera we used was the near-infrared Grism Spectrometer and Imager (GRIM I). The SPIREX telescope was able to enjoy a nearly uninterrupted view of this event from its unique vantage point at the South Pole. We observed during eighteen of the predicted impact times. We carried out two programs. Series of 2.36 and 2.22 micron images were taken to generate light curves for the events. Between impacts, we imaged the disk of Jupiter with most of our filters at hourly intervals. The SPIREX dataset contains nearly 3000 images taken during and after these impacts. We tested the instrumental characteristics and present them here. In addition, we developed tools for automatic data correction and extraction. We present a reduction and analysis of the impact observations including the resulting light curves. Impact timings, energies and phenomenology are tabulated. The light curves are used to investigate properties of the impacts. Most importantly, we have concluded that the impacts of the middle-train fragments differed from those of the other fragments. They produced what we call a prolonged fountain of ejecta. We discuss the significance of this finding. Finally, the onset of CO emission is investigated.

Daniel Reichart

I present the strongest evidence to date for a connection between gamma-ray bursts (GRBs) and supernovae (SNe). In particular, I show that the unusual brightening and reddening of the optical afterglow of GRB 970228 about one month after the burst is naturally explained by a supernova. Furthermore, I present a possible, Cepheid- like luminosity estimator for the long bursts based on the variability of their light curves. The luminosity estimator can be used to measure luminosity distances to the long bursts that are accurate to a factor of ~2. Also, the luminosity estimator suggests that GRB 980425 may well be associated with SN 1998bw at a redshift of z = 0.0085, and that GRB 980425 and the cosmological bursts may share a common physical origin. Additionally, I present a very general, Bayesian inference formalism for testing burst afterglow models and constraining the parameter values of acceptable models, and I present models for how extinction by dust, both in host galaxies and in our galaxy, and absorption by the Lyα forest and by H I in host galaxies, change the intrinsic spectrum of afterglows. Finally, I apply this Bayesian inference formalism and these extinction and absorption models to the measurements of the afterglow and host galaxy of GRB 971214.

Michael Zingale

We present the results of a numerical study of helium detonations on the surfaces of neutron stars. We describe two dimensional simulations of the evolution of a detonation as it breaks through the accreted envelope of the neutron star and propagates laterally through the accreted material. The detonation front propagates laterally at nearly the Chapman-Jouguet velocity, v = 1.3 × 109 cm s-1. A series of surface waves propagate across the pool of hot ash behind the detonation front with the same speed, matching the speed expected from shallow water wave theory. The entire envelope oscillates in the gravitational potential well of the neutron star with a period of ~50 μs. The photosphere reaches an estimated height of 10 km above the surface of the neutron star. Our study confirms that such a detonation can insure the spread of burning over the entire neutron star surface on a timescale consistent with burst rise times. We analyze the sensitivity of the results to the spatial resolution and the assumed initial conditions. We conclude by presenting a comparison of this model to Type I X-ray bursts.

Aparna Venkatesan

Current data of high-redshift absorption-line systems imply that universal reionization of hydrogen occurred before redshifts of about 5. Previous works on reionization by the first stars or quasars have shown that such scenarios are described by a large number of cosmological and astrophysical parameters. Here, we adopt a semi-analytic model of stellar reionization in order to quantify how the optical depth to reionization depends on such parameters, and combine this with constraints from the cosmic microwave background (CMB). We find this approach to be particularly useful in alleviating the well-known degeneracy in CMB parameter extraction between the optical depth to reionization and the amplitude of the primordial power spectrum, due to the complementary information from the reionization model. We also examine translating independent limits on astrophysical parameters into those on cosmological parameters, or conversely, how improved determinations of cosmological parameters will constrain astrophysical unknowns.