Rollins, RP; (2015) Chemical and statistical models of the interstellar medium and star-forming regions. Doctoral thesis , UCL (University College London).

Text thesis-rrollins.pdf Available under License : See the attached licence file. Download (5MB)

Abstract

Astrochemistry is a field currently in a data–rich era, but observations alone cannot provide a complete description of the objects that we see; sophisticated modelling techniques are needed to extract the maximum astrophysical understanding from the data. Chemical models of interstellar gas and ices are continually being developed to provide more accurate descriptions of the chemical composition of our universe. Furthermore, Bayesian inference and statistical learning methods are starting to be incorporated into the analysis of both observational data and astrochemical models. These are the motivations for this thesis. We develop a Markov chain Monte Carlo Bayesian statistics code based on nested sampling for parameter estimation and model comparison with optimal parallel performance. We assess its performance for a set of test problems in terms of accuracy, reliability, time–to–solution and sampling efficiency on two of the country’s top supercomputers. The soft- ware is applied to a chemical model for diffuse clouds in the interstellar medium. Properties of the gas such as its density and temperature are estimated from chemical observations in a statistically robust manner which enables us to comment on the importance of various chemical mechanisms such as photochemistry and grain–surface chemistry in controlling the chemical composition. We also present a study of the effects of mutual shielding by atomic carbon in promoting nitrogen chemistry. By solving the radiative transfer we are able to show that the carbon continuum can have a chemically significant effect on the photoionization and photodissociation rates of a number of species including molecular nitrogen and CN. Finally, we develop a chemical model for carbon monoxide and HCO+ in protostellar outflows where hot outflowing material mixes turbulently and reacts with the cold molecular envelope, allowing us to address an apparent bias towards outflows with large opening angles.

Download activity - last month

Export as JSONCSVXML

Download activity - last 12 months

Export as JSONXMLCSV

Downloads by country - last 12 months

Export as JSONXMLCSV

Archive Staff Only

View Item