Extending and Implementing the Self-Calibration of Intrinsic Alignments of Galaxies to Current and Future Weak Lensing Surveys





Journal Title

Journal ISSN

Volume Title




This dissertation is about extending and applying the self-calibration method for mitigation of intrinsic alignments (IAs) of galaxies. These IA correlations are due to galaxy formation earlier processes that act as a systematic effect to weak gravitational lensing (WL). For ongoing and future WL surveys to reach their full potential as cosmological probes, these IA systematics must be isolated and removed.

There are two types of IAs, the “gravitational shear - intrinsic ellipticity” (GI) and the “intrinsic ellipticity - intrinsic ellipticity” (II) correlation. It can affect the lensing power spectrum at 10% level and cause a mis-estimation of the cosmological parameters, especially in the dark energy equation of state where the bias can be up to 50% and the amplitude of matter power spectrum can be up to 30% . The Self-Calibration (SC) technique has been introduced to subtract the shear-intrinsic (GI) IA contamination in a photometric redshift (photo-z) survey, at a level of 90%. In this work, we present a comparison between the SC technique and the traditional method of modeling IA for an LSST-like survey. We also consider the effect of photometric redshift (photo-z). By comparing the confidence contours of applying SC and using marginalization with modeling IA, we show that SC is competitive with modeling IA method and is applicable to data.

We show how several aspects of SC can be improved. The efficiency of the SC method is high in two redshift bins at large separation, but is relatively low in close bin-pairs. Also it doesn’t apply to the auto-spectra where the (II) signal is significant. We present a novel SC method that can solve these problems. By combining our new SC method with the previously existing SC, the effciency for the GI signal in adjacent bins can be improved from ∼ 90% to ∼ 97%, and the II can be measured with ∼ 1% level bias in all bin-pairs. This combined method can largely reduce the residual bias of the SC method and hence improve the accuracy of the best-fit cosmological parameters. We further summarized the current work on SC and discussed its assumptions on the IA models.

Finally, we apply the SC to KiDS-450 survey data to test the applicability of the SC method. This photometric survey of ∼ 1.5×107 galaxies is one of the best currently available samples to use. Despite some well-known issues with the photo-zs in this sample, we find what appears to be a significant detection of IA using SC. However, due to the photo-z unreliability, this can be considered as a proof of principle for this moment. Future applications of SC to incoming and future data from surveys such as DES and LSST will allow one to obtain more decisive results using SC.



Cosmology, Galaxies—Formation, Calibration, Dark energy (Astronomy)



©2018 The Author. Digital access to this material is made possible by the Eugene McDermott Library. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.