Ph.D Thesis

The Impact of Massive Starbursts on the Chemical Evolution of Galaxies

by Henry A. Kobulnicky

Under the supervision of Professor Evan D. Skillman

ABSTRACT

Young, compact star clusters containing hundred to thousands of the most massive OB and Wolf-Rayet type stars are common features of actively star-forming galaxies. Radio-wave HI and millimeter-wave CO aperture synthesis observations of the interstellar gas in several such systems reveal strong evidence for recent collisions or mergers with other galaxies which probably triggered the present burst. Owens Valley Millimeter Interferometer observations of the proto-typical Wolf-Rayet Galaxy Henize 2-10 are the first ever reported of a Blue Compact galaxy, and they show a distinct dynamical Co feature of ~10^6 solar masses - direct evidence for the interaction hypothesis. Very Large Array neutral hydrogen observations of NGC 5253 show an unusual velocity field that indicates most of the HI is rotating or infalling along the optical axis of the galaxy.
 
Because their massive stars provide the majority of the optical luminosity, ionizing photons, and mechanical energy input to the surrounding medium, massive starclusters are a dominant factor in the chemical and dynamical evolution of galaxies. Most of the oxygen in the universe, and to a lesser extent carbon and nitrogen, is synthesized with in massive stars and returned to the interstellar gas via strong stellar winds and supernova explosions as these stars complete their evolution. Yet, spatially-resolved optical and ultraviolet spectroscopic investigations of the ionized gas in several starburst galaxies fail to find any sing of recent nucleosynthesis products in the vicinity of evolved starclusters. The chemical abundances of O, N, He and probably C, appear very homogeneous on scales of ~ 1 kpc or less, despite the fact that models of the chemical enrichment expected from a single 10^6 solar mass burst show that large localized chemical enhancements should occur. That there is no evidence of localized chemical enrichment within the HII regions of most metal-poor galaxies suggests the recently-released heavy elements are 'hiding' either in a hot, 10^6 phase or in a cool neutral atomic or molecular phase. In either case, the timescale for visible enrichment in galaxies appears to exceed the lifetimes of the HII regions and the spatial scales must exceed 1 kpc. These data are inconsistent with the HII region "self-enrichment" or "pollution" hypothesis. For the moment, the heavy elements produced in starbursts can be considered "mission", but upcoming X-ray and IR observatories may be able to establish their physical phase and location.
 
The use of the C+/O+ ratio is developed as a diagnostic of the C/O ration in HII regions. Because O,C, and N content of the stars and gas within galaxies contains a fossil record of previous star formation activity, measurements of the relative abundances of these elements provide clues about stars which have long since disappeared. The C/O and N/O ratios, in particular, may be used as a clock to measure the time interval since the last major episode of star formation. Hubble Space Telescope spectroscopic measurements show evidence for a correlation between C and N abundances among galaxies with similar metallicity (O/H). The existence of such a correlation implies that C and N production mechanisms are coupled. This is consistent with the hypothesis that both the N and C abundances are the result, not of local, temporal enrichment from massive stars, but global, secular enrichment resulting from the particular star formation history of low and intermediate mass stars in each object.
 
 

Master's Thesis

R-Band Polarimetry of Cygnus OB2: Implications for the Magnetic Field Geometry and Polarization Models

by Henry A. Kobulnicky

Under the supervision of Professor Terry Jones

ABSTRACT

 

We present new r-band polarimetry of 132 members of the Cygnus OB2 association. From these data we have determined the angular coherence length of the polarization position angle and tested two models for interstellar polarization toward a region where the Galactic magnetic field is expected to be primarily along the line of sight. The polarization magnitudes and position angles in our sample decorrelate on size scales of ~ 0.5°, compared to 0.2° for the optical extinction. This constrains the variations in the magnetic field orientation to sizes less than 8 pc. The distribution of polarization magnitudes and position angles is in good agreement with a model which describes the polarization in terms of preferential extinction in small dark clouds threaded by a two-component magnetic field. One component is constant in direction, while the second is random in orientation and is characterized by the amplitude of Alfven waves in the interstellar medium. A similar model, in which the magnitude of the random component has a Gaussian distribution and an orientation evenly distributed through 4 pi steradians, is also consistent with the data. We explore the predictions of each model for the polarization magnitude and position angle as a function of the orientation of the constant component and point out several key differences that might be exploited to further test their potential for describing interstellar polarization.
We find the polarization position angles of stars in the vicinity of Cygnus X-3 are approximately orthogonal to the position angle of the elliptical radio scattering reported by Molnar et al. (1989). If the scattering is taking place on the near side of Cyg OB2, the data are consistent with mechanisms that require the major axis of the scattering ellipse to lie perpendicular to the projected magnetic field.