Ph.D. Thesis Defenses: 2001
Detection and analysis of the tidal tails around the globular cluster Palomar 5 in the SDSS commissioning data
May 10, 2001 | PhD Advisor: Donald G. York | PhD Thesis Defense
Constance Rockosi

The tidal tails around the globular cluster Palomar 5 are analyzed over a 41deg2 area of the SDSS photometric catalogs. The matched filter algorithm provides the maximum possible signal-to-noise detection of the cluster stars over the measured background, and the expected and actual effectiveness of the technique in the context of this dataset is discussed. The stellar background is examined in some detail for systematic variation as a function of Galactic position in order to assess its effect on the detection efficiency. Of the total number of Pal 5 stars detected 45% are out in the tails, which when the differing tidal radii are accounted for is in close agreement with the fraction found by Odenkirchen et al. (2001a). The tails are found as the only additional 3sigma overdensity of cluster stars over the entire 41deg2 area studied. The annular-averaged density of stars along the tails is fit to a power-law in radius with best-fit index -1.58, significantly steeper than that predicted from a constant orbit-averaged mass-loss rate.

The far-infrared/submillimeter polarization spectrum of molecular clouds and analysis based on temperature maps of Orion
May 24, 2001 | PhD Advisor: Roger H. Hildebrand | PhD Thesis Defense
John Vaillancourt

We compile polarization data on galactic molecular clouds from instruments spanning the wavelength range 60-1300 μm. We show that the polarization spectrum in these clouds falls from 60-350 μm but rises from 350-1300 μm. To explain this spectrum we require a model in which the emission arises from dust grains at multiple temperatures along the line of sight and where the polarizing efficiency of these grains is correlated with their temperature. In order to test this hypothesis we collect flux data from the literature and create spectral energy distributions (SEDs) from 38-1100 μm in a ~5' x 5' region of the Orion A molecular cloud. Dust temperature distributions are estimated from these SEDs on a point- by-point basis within the cloud at ~30' resolution. We find cold (20 K), dense cores associated with submillimeter flux peaks and warm (50 K) dust associated with the M42 H II region. The SEDs are well fit by both one- and two-temperature components in the dust temperature distribution. This is due to the inadequate quality of the flux data and the sparsely sampled spectral points. While the results are consistent with our multiple temperature model they do not provide a conclusive test. However, we show that this problem should be resolved with better data from SOFIA.

Layer formation in semiconvection
June 26, 2001 | PhD Advisor: Robert Rosner | PhD Thesis Defense
Joseph Biello

Layer formation in a thermally destabilized fluid with stable density gradient has been observed in laboratory experiments and has been proposed as a mechanism for mixing molecular weight in late stages of stellar evolution in regions which are unstable to semiconvection. It is not yet known whether such layers can exist in a very low viscosity fluid: this work undertakes to address that question. Layering is simulated numerically both at high Prandtl number (relevant to the laboratory) in order to describe the onset of layering instability, and the astrophysically important case of low Prandtl number. It is argued that the critical stability parameter for interfaces between layers, the Richardson number, increases with decreasing Prandtl number. Throughout the simulations the fluid has a tendency to form large scale flows in the first convecting layer, but only at low Prandtl number do such structures have dramatic consequences for layering. These flows are shown to drive large interfacial waves whose breaking contributes to significant mixing across the interface. An effective diffusion coefficient is determined from the simulation and is shown to be much greater than the predictions of both an enhanced diffusion model and one which specifically incorporates wave breaking. The results further suggest that molecular weight gradient interfaces are ineffective barriers to mixing even when specified as initial conditions, such as would arise when a compositional gradient is redistributed by another mechanism than buoyancy, such as rotation or internal waves.

Determining the cosmic distance scale from interferometric measurements of the Sunyaev-Zel'dovich effect
June 28, 2001 | PhD Advisor: John E. Carlstrom | PhD Thesis Defense
Erik Reese

We determine the distances to 18 galaxy clusters ranging from z ~ 0.14 to z ~ 0.78 from a maximum likelihood joint analysis of 30 GHz interferometric Sunyaev-Zel'dovich effect (SZE) and X-ray observations. We model the intracluster medium (ICM) using a spherical isothermal β model. We quantify the statistical and systematic uncertainties inherent to these direct distance measurements, and we determine constraints on the Hubble parameter for three different cosmologies. These distances imply a Hubble constant of 60+4+14-4 -19 km s-1 Mpc-1 for an Ω M = 0.3, ΩΛ = 0.7 cosmology, where the uncertainties correspond to statistical followed by systematic at 68% confidence. The best fit H0 is 56 km s -1 Mpc-1 for an open Ω M = 0.3 universe and 54 km s-1 Mpc -1 for a flat ΩM = 1 universe. With a sample of 18 clusters, systematic uncertainties clearly dominate. The systematics are observationally approachable and will be addressed in the coming years through the current generation of X-ray satellites (Chandra & XMM-Newton) and available radio observatories (OVRO, BIMA, & VLA). Analysis of high redshift clusters detected in future SZE and X-ray surveys will allow a determination of the geometry of the universe from SZE determined distances, providing a distance ladder independent check on recent constraints on the cosmology of the universe.

Spectroscopy of trihydrogen(+) in laboratory and astrophysical plasmas
July 2, 2001 | PhD Advisor: Takeshi Oka | PhD Thesis Defense
Benjamin McCall

H+3 is the simplest and most fundamental polyatomic molecule, consisting of only three protons and two electrons. As such, it plays important roles in the laboratory spectroscopy of hydrogen-rich plasmas, the theoretical calculation of rotation-vibration energy levels, and also the chemistry of interstellar clouds. This dissertation touches on all three of these areas. High resolution spectroscopy of H+3 has been performed in a positive column discharge. The combination bands ν1 + 2ν2 <-- 0 and 2ν1 + ν2 <-- 0 have been observed with a diode laser, and reach the highest energy vibrational states yet studied. The initial detection of the fourth overtone band (5ν2 <-- 0, which reaches above the barrier to linearity) with a Titanium-Sapphire laser is also discussed. A comprehensive re-evaluation of all previous laboratory spectroscopy of H+3 has been conducted in order to obtain a reliable linelist and derive experimentally determined energy levels. These energy levels have been compared with the most recent variational calculations on ab initio potential energy surfaces. It is hoped that this comparison will permit further refinements in the theoretical calculations. H+3 has been detected (using absorption lines of the ν2 fundamental band) in several dense interstellar clouds, where it serves as the universal protonator, initiating a chain of ion-neutral reactions that is responsible for the production of the variety of molecules observed by radioastronomers. In dense clouds, measurements of H+3 provide direct estimates of the clouds' path lengths, average number densities, and kinetic temperatures. H+3 has also been observed in several diffuse interstellar clouds, where it is supposed to be two to three orders of magnitude less abundant due to the efficiency of electron recombination. This observational result suggests a serious general problem with the models of diffuse cloud chemistry. The most likely solution is that the ratio of the cosmic ray ionization rate (ζ) to the dissociative recombination rate (ke) is at least one to two orders of magnitude higher than has been generally assumed. This implies that the laboratory measurement of ke is not applicable to interstellar conditions, and/or that H2 ionization is enhanced in diffuse clouds relative to dense clouds.

Galaxy clusters and cosmology with the Sunyaev-Zel'dovich effect and weak lensing
July 11, 2001 | PhD Advisor: Michael S. Turner | PhD Thesis Defense
Gilbert Holder

Sunyaev-Zel'dovich effect (SZE) measurements and weak lensing observations are well-matched in that both probe the projected mass of galaxy clusters. When gas mass fractions derived from SZE-weak lensing comparisons are combined with the hydrostatic method of estimating gas fractions, a direct estimate of the temperature can be made, in a manner that is nearly independent of cosmology. Further assuming a calibrated measure of the gas fraction allows direct distance determinations using only SZE and weak lensing data, without X-ray information. SZE and weak lensing observations are also well matched in terms of the filters that both are applying to the underlying mass distributions. In this work we explicitly show this connection and detail and validate a non-parametric method for calculating gas fractions by combining SZE measurements and weak lensing observations, using the example of interferometric SZE data. The method consists of a comparison of Fourier modes at the spatial scales that have been measured in both the SZE and weak lensing data. We demonstrate the robustness of the method by generating mass maps and SZE images of galaxy clusters from N-body+gas simulations and realistic interferometric observation strategies. We show that current observations should be able to constrain the gas fraction to ˜20% for clusters at redshifts between 0.1 and 1, depending on the accuracy to which the gas temperature and redshift distribution of lensed sources are known. Current data is of a quality which should allow individual determinations of the angular diameter distance to an accuracy of ˜30%.