Le contenu de cette page n’est pas disponible en français. Veuillez nous en excuser.
 

From precision to accuracy: cosmology with large imaging surveys



Playing this video requires the latest flash player from Adobe.

Download link (right click and 'save-as') for playing in VLC or other compatible player.


Recording Details

Speaker(s): 
Scientific Areas: 
Collection/Series: 
PIRSA Number: 
15040046

Abstract

Photometric surveys are often larger and extend to fainter magnitudes than spectroscopic samples, and can therefore yield more precise cosmological measurements. However, photometric data are significantly contaminated by multiple sources of systematics, either intrinsic, observational, or instrumental. These systematics affect the properties of the raw images in complex ways, propagate into the final catalogues, and create spurious spatial correlations. Some of these correlations may also be imprinted in spectroscopic catalogues, since the latter rely on targets selected from imaging data. Therefore, not just precise — but also accurate — cosmological inferences from imaging surveys require careful mitigation of spatially-varying systematics. I will present a new framework of extended mode projection to robustly mitigate the impact of such systematics on power spectrum measurements. I will demonstrate the effectiveness of the technique, showing constraints on primordial non-Gaussianity using the clustering of 800,000 photometric quasars from the Sloan Digital Sky Survey in the redshift range 0.5 < z < 3.5. Finally, I will present a framework to map the observing conditions of the Science Verification data in the Dark Energy Survey (DES) and incorporate them into end-to-end simulations of the DES transfer function. The considerations presented here are relevant to all multi-epoch surveys, and will be essential for exploiting future high-cadence surveys such as the Large Synoptic Survey Telescope (LSST), which will require detailed null-tests and realistic end-to-end image simulations for correct cosmological inferences.