SKA weak lensing – II. Simulated performance and survey design considerations

Anna Bonaldi, Ian Harrison, Stefano Camera, Michael Brown

    Research output: Contribution to journalArticlepeer-review

    110 Downloads (Pure)

    Abstract

    We construct a pipeline for simulating weak lensing cosmology surveys with the Square Kilometre Array (SKA), taking as inputs telescope sensitivity curves; correlated source flux, size and redshift distributions; a simple ionospheric model; source redshift and ellipticity measurement errors. We then use this simulation pipeline to optimize a 2-yr weak lensing survey performed with the first deployment of the SKA (SKA1). Our assessments are based on the total signal to noise of the recovered shear power spectra, a metric that we find to correlate very well with a standard dark energy figure of merit. We first consider the choice of frequency band, trading off increases in number counts at lower frequencies against poorer resolution; our analysis strongly prefers the higher frequency Band 2 (950–1760 MHz) channel of the SKA-MID telescope to the lower frequency Band 1 (350–1050 MHz). Best results would be obtained by allowing the centre of Band 2 to shift towards lower frequency, around 1.1 GHz. We then move on to consider survey size, finding that an area of 5000 deg2 is optimal for most SKA1 instrumental configurations. Finally, we forecast the performance of a weak lensing survey with the second deployment of the SKA. The increased survey size (3π steradian) and sensitivity improves both the signal to noise and the dark energy metrics by two orders of magnitude.
    Original languageEnglish
    Pages (from-to)3686-3698
    JournalMonthly Notices of the Royal Astronomical Society
    Volume463
    Issue number4
    Early online date24 Aug 2016
    DOIs
    Publication statusPublished - Dec 2016

    Keywords

    • Dark energy
    • Dark matter
    • Large-scale structure of the Universe

    Fingerprint

    Dive into the research topics of 'SKA weak lensing – II. Simulated performance and survey design considerations'. Together they form a unique fingerprint.

    Cite this