Benjamin Giblin

In this work, we analyse the latest data-release from the Kilo-Degree Survey (KiDS). KiDS has used the VLT Survey Telescope to observe 1006 square degrees of the sky in multiple optical bands. The subsequent data set, labelled 'KiDS-1000', contains measurements of the distortions in the shapes of 21 million galaxies. These distortions are induced by the galaxies being gravitationally lensed by the matter in our Universe which resides between said galaxies and our telescopes on Earth. The galaxy shape distortions, or 'shear', along with the galaxy redshifts, are used to place constraints on cosmological parameters, namely the matter energy density and the amplitude of matter's clustering. In Figure 1 we show maps of the projected matter density in the Universe, inferred from the KiDS-1000 data set.

We proceed to check that the accuracy of our modelling of the 'point-spread function' (PSF). The PSF models the galaxy shape distortions induced by effects other than gravitational lensing, e.g., from Earth's atmosphere or telescope/camera properties. The PSF pattern across the pixels in our camera, and expressed in terms of two ellipticity components, is shown in the upper panels of Figure 2. The featureless residual pattern after subtracting our PSF model, shown in the lower panels, are a sign that our model is fairly accurate. We additionally test prescriptions for how imperfect PSF modelling biases our measurement of the 'shear correlation function' - a statistic which measure the correlations in galaxy shapes as a function of galaxy separation, and is used in the likelihood to place constraints on cosmological parameters. The systematic contributions to this measurement are shown by the 3 coloured lines in Figure 3 and are found to have very small impact on our cosmological parameter constraints.

Lastly, we measure the gravitational lensing signal of the KiDS-1000 galaxies around the positions of foreground galaxies, referred to as 'tangential shear'. By comparing the 2D (Cartesian) tangential shear pattern around foreground galaxies (upper panel of Figure 4) to the 1D (angular) tangential shear pattern (middle panel), we can detect if there is an additive bias in our shear measurement method. The featureless residuals between the 2D and 1D signals (lower panel of Figure 4) implies no such bias is present in our data. In our final test, we measure the 1D angular tangential shear, binned by the foreground and background galaxy redshifts. This is shown by the blue points in the various panels of Figure 5, each of which corresponds to a different redshift bin combination. The consistency of this measurement with the theoretical prediction, shown by the red lines, is a sign that the redshifts estimated for our background galaxies are accurate.

In conclusion, we find that the KiDS-1000 data set passes all of the tests to validate the shear and redshift measurement methods. This means that this data is scientifically robust and ready to be used in analyses to constrain the cosmological parameters of our Universe.