Copy the page URI to the clipboard
Weiner, Charles Frederick
(2019).
DOI: https://doi.org/10.21954/ou.ro.00010bcb
Abstract
The objective of this project is to examine the extent to which strong gravitational lensing can constrain cosmological parameters. I present the results of applying a modified version of an existing model of strong gravitational lensing to forthcoming surveys by Euclid, currently scheduled to launch in 2021, and investigate also how the model may be adapted further to accommodate background galaxy sources that have not previously been included. The initial model, on which the modifications are based, was first constructed by Dr Tom Collett (Collett 2015) with the source code, at the time of writing, freely available on https://github.com/tcollett/LensPop. The study commences with a review of the existing model's code, which includes a mapping of the key dependencies. As a natural consequence of this, discrepancies that I have identified within the code are detailed, as are inconsistencies with the supporting article by Collett (2015) in which the principal features of his model are described. Once the discrepancies are corrected, or otherwise resolved, the modified model is run and the implications of these assessed: most are found to be minor, although more significant issues arise when the model is tested under non-standard cosmologies. An analysis of the results for both Euclid's Wide Field and Deep Field surveys is presented using the modified model, as are predictions by the model for the forthcoming Cosmic Evolution Survey (COSMOS) and the Wide Field InfraRed Survey Telescope (WFIRST). In comparison to the Wide Field survey, the model's prediction of a 7-fold increase in the sky density of detectable lenses for the Euclid Deep Field survey is found to be mainly due to an increased sensitivity of 2 magnitudes in the latter. For the COSMOS survey, a prediction of some 120 lenses suggests that further lensing systems have yet to be confirmed in the survey field whereas, in the case of WFIRST, a prediction of just under 100,000 lenses means its increased depth almost compensates for the smaller area, when compared to the Euclid Wide Field survey; compared to the Euclid Deep Field survey, on the other hand, WFIRST is both wider and deeper, with this prediction representing a 25-fold increase in the number of discoverable lenses. The extent to which the model can constrain cosmological parameters is then considered. This requires an investigation not only of the model's direct sensitivity to a cosmology by virtue of the lensing equations, but also of the model's sensitivity to any astrophysical assumptions, such as those governing density or luminosity functions, that are intrinsic to the code. I find there is prima facie evidence that the model does constrain the cosmologies tested, and conclude also that it is not particularly sensitive to those astrophysical assumptions. Finally, by replacing the simulated source data described in Collett (2015) with a more appropriate mock catalogue, I examine the predictions of the model when submillimetre galaxies are considered. In this respect, a source population comprising solely submillimetre galaxies gives rise to an under-prediction by the model of the number of lenses, when compared to other studies; furthermore, once adapted in this fashion, the model does not impose any significant constraints on the cosmologies tested.
CORE (COnnecting REpositories)
CORE (COnnecting REpositories)
