Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.799322
Title: Finding and looking through strong gravitational lenses
Author: Hartley, Philippa
ISNI:       0000 0004 8504 2997
Awarding Body: University of Manchester
Current Institution: University of Manchester
Date of Award: 2019
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
Strong gravitational lensing is a rare but important astrophysical phenomenon. Manifesting as dramatic arcs and multiple objects, the pattern of a lensed background source is dependent on the distribution of intervening matter: both baryonic and dark. For this reason, strong lens systems can be used to probe the invisible dark matter structure of lensing objects, placing constraints on the size of dark matter components in order to test the LambdaCDM paradigm. The magnifying power of lensing also permits greater study of the lensed source; by using lenses as 'cosmic telescopes', observers have access to the very faintest and most distant objects in space. A single lens system can yield a rich body of scientific results concerning the objects involved. A thousand lens systems, however, would provide a statically robust sample which can be used to place very tight constraints on the Hubble constant, on galaxy evolution and even on the nature of dark energy. Imminent large sky surveys will image millions of strong lenses but finding those lenses among the billions of objects observed will be impossible to do by eye. This thesis presents the development of a novel machine learning method which automates the lens finding process. The tool uses a support vector machine to classify lens candidates, and won a prize in the 2017 Strong Lens Finding Challenge. It was subsequently applied to data from the Kilo Degree Survey, finding several highly likely new lens systems. This thesis also presents results from the detailed study of three strongly lensed radio quiet quasars (RQQ). These objects are of great interest not only due to their ability both to trace dark matter substructure and to measure the expansion of space, but also since it is unknown whether their radio emission results from the AGN activity seen in the optical or whether starburst -- or more exotic -- activity is responsible. EVN and e-MERLIN observations of RQQ HS~0810+2554 have revealed brightness temperatures in excess of the limit from starburst emission, and subsequent modelling predicts a linear alignment of two highly compact components with the optical core, providing conclusive evidence for AGN jet emission. VLA observations of PG~1115+080 and J1004+4112 similarly find evidence of AGN activity, in the very long and narrow shape of the modelled source and in the variability of the source, respectively. The work advances the programme to determine once and for all the role of RQQ emission mechanisms in galaxy feedback models, and demonstrates the importance of high quality observations in understanding the phenomena associated with galaxy evolution.
Supervisor: Jackson, Neal ; Battye, Richard Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.799322  DOI: Not available
Keywords: Radio jets ; Artificial intelligence ; Machine learning ; Radio quiet quasars ; Strong gravitational lensing ; Cosmology ; Radio interferometry ; Quasars
Share: