Diagnostics of Evolving Planetary Nebulae: Observation and Simulation of Atomic and Molecular Emission Lines

Abstract

Planetary nebulae (PNe) are the glowing, gaseous remnants of low to intermediate mass stars. As light from the star passes through the layers of gas and dust, a distinct spectrum is produced, with a variety of emission, absorption and continuum features. This means that a large number of physical properties can be inferred from a single spectroscopic observation. This thesis concentrates on the properties of evolving PNe which are bright in the 1-0 S(1) emission line of molecular hydrogen (H2), observed by the UWISH2 survey.

The first part of this thesis is the reduction and analysis of K-band spectra of 29 true and candidate PNe from this survey. The Bry line is identified in all but two objects, signifying the presence of H+. 13 candidate PNe have spectra with the Bry line, but show no signs of Hα emission in optical surveys, confirming the need for infrared observations to reveal PNe obscured by dust in the Galactic Plane. Ratios of fluxes of H2 and hydrogen recombination lines are used to link the dominant H2 excitation mechanism with PN morphology and evolution. In line with previous results, we find that UV-fluorescence plays a greater role in W-BPNe (bipolar PNe with pinched waists), while R-BPNe (bipolar PNe with ring structures) are predominantly thermally excited. Using the relative spatial extents of HII and H2 emission, which act as a proxy for evolution, we find that W-BPNe are likely to be younger objects, while the R-BPNe are more evolved.

The second part focusses on Abell 53 - a little-studied PN consisting of a bright ring of HII and H2 emission. Multi-wavelength observations are obtained, including photometry from the near-UV to radio, along with optical and near-IR long-slit spectra. After correcting for extinction, these observations are used to conduct a detailed analysis in order to determine the properties of Abell 53, including gas temperature and density, along with physical dimensions of radius and distance. The photoionisation code Cloudy is used with the Markov-Chain Monte Carlo (MCMC) optimisation procedure in order to find a best-fit set of parameters to describe the optical observations. A good fit is found using a star approximated by a blackbody with an effective temperature and luminosity of 104 kK and 400 Lʘ respectively, surrounded by a spherical shell spanning from 0.09 to 0.17 pc, with density 830 cm-3 and filling factor 0.48. Observed line ratios of H2 1-0 S(1) to Hα and 2-1 S(1) are reproduced by replacing the outer portion of the previous shell with a higher-density region of ~ 15000 cm-3, thickness 0.010 pc and a constant temperature of 2100 K.

The third part uses a sequence of photoionisation models, such as the one used for Abell 53, in order to replicate H2 emission from an evolving PN. Each model in the sequence represents a PN at a particular evolutionary stage, defined by parameters describing the incident radiation field, expansion and density. Using values representative of a typical PN, Hα surface brightnesses were obtained, comparable to observations and hydrodynamical models, leading us to believe that this method is reliable. The downside of using a sequence of static models, rather than a time-dependent radiation-hydrodynamic model, is that the timescales over which certain reactions take place must not exceed the time intervals between successive models. In order to try to match the surface brightnesses in the UWISH2 survey, constant density models were developed with densities up to the highest observed in knots in PNe. We paid special attention to the reaction rates of H2 formation on grain surfaces, and total H2 destruction, which are highly dependent on density. Unless the density is at the highest end of possibilities, the timescales required to reform H2 on grains is much longer than the intervals between successive models, meaning that the CS flux changes before the models have time to react to the changes. The H2 surface brightness for the 107 cm-3 density model while on the cooling track are of the same order as those in the UWISH2 survey, and so it is likely that this emission is compatible with emission from dense knots.